(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Radio Circuits"

Basic Electronics Series 



RADIO CIRCUITS 



Basic Electronics Series 

RADIO CIRCUITS 

by 

Thomas M. Adams 



With a specially written chapter for 

the guidance of the English reader 

by W. Oliver (G3XT) 



FOULSHAM-SAMS 

TECHNICAL BOOKS 

Published and Distributed by 
W. FOULSHAM & CO. LTD. 

SLOUGH BUCKS ENGLAND 



W. FOULSHAM & CO. LTD. 
Yeovil Road, Slough, Bucks, England. 



RADIO CIRCUITS 

Copyright © 1963 by Howard W. Sams & Co., Inc., 
Indianapolis 6, Indiana. 

Introduction © 1965 by W. Foulsham & Co., Ltd., Slough, 
Bucks, England. 

Reproduction or use, without express permission, of editorial 
or pictorial content, in any manner, is prohibited. No patent 
liability is assumed with respect to the use of the information 
contained herein. 

Library of Congress Catalog Card Number: 63—13054 

Cat. Code No. BER— 1 



Introduction Printed and Made in Great Britain 

by East Midland Printing Company Limited, Bury St. Edmunds. 

Balance printed in U.S.A. 



It is essential that the English reader should read this chapter. 



If you have already studied the four previous volumes in this 
Basic Electronics series, you will be familiar with the colour system 
which forms such an original and helpful feature of the diagrams 
illustrating the text. 

These coloured lines — red, blue, green and so on — are super- 
imposed on the ordinary conventional type of black-and-white 
circuit diagrams (or "schematics" as they are termed in America) 
and serve several useful purposes. 

First, they give a clear visual key to which of the various 
components and connections belong to each respective part of the 
circuit. You can thus analyse the whole arrangement at a glance, 
separating the signal-paths comprising a grid circuit, for example, 
from those forming an anode circuit; and so on. 

Second, the arrows on the coloured lines give an idea of the 
theoretical direction of current-flow involved in the functioning of 
any particular circuit. 

Third, a comparison of the related diagrams for each circuit shows 
the movements and changes which take place at each cycle of 
operation. 

Whereas the previous books in this Basic Electronic Series were 
concerned with circuits falling into certain broad classifications — 
e.g., Amplifier Circuits, Detector Circuits, Transistor Circuits, and 
Oscillator — the present volume brings these together and deals with 
three typical kinds of complete receiver— the superhet, the transistor 
and the tuned-radio-frequency types. 

The first five chapters are concerned with individual features or 
functions of these typical receivers, such as heterodyning actions 
(which occur in mixer and converter arrangements, as well as in a 
BFO unit), noise-limiters, reaction or regenerative devices, AVC 
circuits and a typical power-supply pack using a half-wave rectifier. 

The discussions are widened from the theoretical, "how it works" 
angle to the practical "trouble-shooting" angle so that the details of 
current flow in the various diagrams can be used not only for 
informative purposes but also as a practical aid to fault-tracing. 

In other words, the diagrams show what currents should be 
present in any given part of the circuit; and test procedures are 
described in the text which will enable you to confirm whether or 
not these currents are there. From the presence or absence of these 
currents one can deduce which parts are faulty and form a fairly 
accurate idea of what is wrong with them. 

Some of the diagrams have been simplified to a very considerable 



extent; and therefore the beginner should bear in mind that the 
practical circuit of an actual, specific receiver may well be a good 
deal more complex than one might suppose from a casual glance at 
the somewhat skeletonized or partial diagrams in this book. 

For example, an IF transformer may be shown, for the sake of 
simplicity, as consisting merely of a pair of coil-windings (primary 
and secondary) with a fixed capacity across the primary and none 
at all across the secondary; and with no means of adjusting the 
effective value of inductance. 

In a practical receiver, however, there would be full provision 
for peaking or aligning the tuning of both primary and secondary 
windings to the chosen intermediate frequency. 

The method of doing this varies in different sets; for example, 
the arrangements may comprise a pair of semi-variable (pre-set) 
capacitors, one across each winding. Or, again, there may be adjust- 
able iron-dust cores to the windings themselves. The former method 
was popular in earlier sets; but the latter is usual practice nowadays. 

Since this book is of American origin, the only items likely to 
need comment or explanation for the benefit of British readers are 
the essential points of difference between British and American 
electronic technique, mains voltage, etc. 

Regarding intermediate frequency, the author quotes 455 kc as 
being the usual in most sets. This is a favourite frequency for 
American sets; but in Britain many different frequencies have been 
used in radio receivers by various makers. 455 is among them, but 
465 is more usual in this country, being also a popular figure for IF 
transformers designed for the home constructor. 

The following are some general notes, relating not only to matters 
relevant to the present book but also to other points of difference 
which may need explanation if you are studying other technical 
literature originating from the United States and dealing with 
various branches of electronics. 



IMPORTANT DIFFERENCES BETWEEN BRITISH 
AND AMERICAN ELECTRICITY AND ELECTRONICS 

There are important differences between British and American 
electricity and electronics which must always be heeded. From the 
general survey of these given below you can select items which fit 
any particular case. 

If in doubt, especially regarding safety, be sure to seek expert 
advice from a local source, such as your nearest electricity show- 
room-centre or a competent radio, television or electrical engineer. 



Any misuse of mains apparatus can be dangerous; never take 
unnecessary risks. 

MAINS. British voltage is usually twice as high as American; 
and much more dangerous. Never forget this. American domestic 
appliances, radio sets and the like, are normally designed for 110- 
120 volts AC at 60 cycles per second. British electric mains are 
mostly rated now at 230-240 volts AC at 50 cycles per second. But 
there are exceptions; so check locally. 

American apparatus and designs have to be converted to suit our 
mains; or a means must be employed to drop our voltage to their 
rating. 

The latter alternative can consist of a heavy-duty mains trans- 
former with standard British primary winding to match our mains 
and a secondary giving an adequate current output at American 
voltage (110-120). This virtually provides American mains facilities 
in British premises. It enables American apparatus to be used with- 
out internal alteration. 

The other choice is to convert each individual set or apparatus 
internally to suit our mains. This is usually the best plan if you are 
building sets, etc., to American designs. 

Suppose, for example, you are building a radio receiver which, in 
the American design, has a mains transformer with 110-120 volt 
primary; and H.T. and L.T. secondaries to suit the valves. All you 
have to do is to substitute an equivalent type of transformer with a 
standard British primary and the same secondary voltages as in the 
American version. The primary may have variable tappings from 
200 to 250 volts and you adjust the voltage-selector to suit your 
local voltage. 

Other methods of voltage-dropping include an auto-transformer; 
mains dropping resistor; line-cord resistor (seldom used nowadays); 
lamps (to dissipate the surplus voltage) or barretters; and a capacity 
method. 

Remember that mains transformers are for AC only. They cannot 
and must not be used on DC mains. For the latter, a mains-dropper 
in the form of a suitable heavy-duty resistance is the best plan. But 
resistors run hot and must be mounted away from inflammable 
materials as well as being given ample ventilation. 

Mains dropping resistance values must be calculated exactly to 
suit the individual circuit. The voltage is dropped at a rate of one 
volt per milliampere of current per thousand ohms of resistance. 
The calculation is based on Ohm's Law; and if you are in any doubt 
get a local technician to check your reckoning before connecting up 
the set. An error can result in expensive damage (to valve 
filaments, etc.). 



The difference of 10 cycles- per second in frequency between 
British and American mains is important in the case of record- 
players and electric clocks where the speed of the motor is actually 
governed by the frequency of the mains. But it can be disregarded 
in ordinary radio sets and so on. 

TECHNICAL TERMS. Certain American technical terms differ 
from ours. The following are typical examples. 

Aerial = Antenna; Earth = Ground; Gramo= Phono (phono- 
graph); Valve = Tube or vacuum-tube. Circuit diagrams are termed 
schematics. Low tension, high tension and grid bias supplies are 
known by the letters A. B and C respectively. 

OFFICIAL REGULATIONS. All information on official 
regulations and licencing rules given in American technical writings 
should be disregarded on this side of the Atlantic, as our rules differ 
in many respects from those in force in the States. 

The American authority dealing with these matters is the FCC 
(Federal Communications Commission). Ours is the GPO. The 
section dealing with amateur transmitting licences is the Amateur 
Radio Licensing Department, GPO Headquarters, St. Martin's le 
Grand, London, E.C.I. 

Anyone interested in short-wave listening or amateur transmitting 
will benefit from becoming a member of the Radio Society of Great 
Britain, New Ruskin House, Little Russell St., London, W.C.I. 
Among its various publications, the Society issues a call-book listing 
British amateur transmitting stations, also a monthly Bulletin (free 
to members). 

The unauthorized or unlicensed use of a transmitter is a punish- 
able offence in this country; and so, of course is the use of a receiver 
without a licence from the G.P.O. 



PREFACE 



This book is presented as a logical follow up to the four pre- 
viously published volumes in this series. The circuits analyzed 
are the basic types widely used in various amplitude-modulated 
receiver systems. The analytical technique is the same as in the 
previous texts — namely, to make a positive identification of each 
and every electron current flowing in the circuit under considera- 
tion, and then to thoroughly describe the movements of that cur- 
rent so that its underlying function or purpose will be made clear. 
In this, as in the other volumes, electron currents have been 
treated as if they were actually moving parts in a piece of me- 
chanical machinery; when this approach is used, it is much easier 
to understand the relationship of each current to every other 
current. 

Individual circuits are discussed in the first five chapters. The 
last three chapters present analyses of a table model superhetero- 
dyne radio, a transistorized superheterodyne radio, and a TRF re- 
ceiver. Standard maintenance tests — voltage checking and signal 
substitution — have been discussed in detail. These tests are nor- 
mally performed by the technician or maintenance man; how- 
ever, an understanding of the significance of the various steps in 
these procedures can teach both student technicians and engi- 
neering students about what happens inside circuits. Voltage 
checking and signal substitution enable the technician to make a 
judgment on almost every component in the radio by observing 
something about one or more of the predicted electron currents 
at each of the many test points. Normally the presence or absence 
of a particular current, or a substantial change in its value, will 
enable the student to determine what component is at fault. 

Occasionally, a circuit component carries several electron cur- 
rents simultaneously; however, the majority of components will 
be found to carry one or, at the most, two separate electron cur- 



rents during normal operation. The circuit diagrams in Chapters 
6 and 7 enable you to "envision" each of these currents. The test 
procedures then give you a means of knowing whether the 
expected current is actually present. Any deviation from the 
expected amount of a particular current can lead to a usually 
obvious deduction as to which component is not performing 
properly. 

Every attempt has been made to keep the discussions as simple 
as possible; little or no background in electronics or mathematics 
is necessary to understand the text material. Because the basic 
electron current approach used is necessary to students of all 
levels in electronics, this book is considered neither too ad- 
vanced for high school and technical institute training, nor too 
elementary for college level. 

Thomas M. Adams 



TABLE OF CONTENTS 



CHAPTER 1 

Heterodyning Actions 7 

Triode Mixer — Pentagrid Mixer Circuit — Pentagrid Converter Cir- 
cuit — Beat-Frequency Oscillator 



CHAPTER 2 

Signal Demodulation and Automatic Control 

of Volume 35 

Detector, AVC, and Audio Amplifier— Generation of Positive AVC 
Voltages 



CHAPTER 3 

Noise Limiting Principles 61 

Shunt-Diode Noise Limiter — Series Diode Noise Limiter — Dual- 
Diode Noise Limiter — Squelch Circuit 



CHAPTER 4 

Half- Wave Power Supply 82 

Half- Wave Rectifier Circuit 



CHAPTER 5 

Regeneration 89 

Regenerative Detector — Superregenerative Receiver 



CHAPTER 6 

Typical Superheterodyne Receiver 104 

IF Amplifier — Audio Power Amplifier— Voltage Checking the Su- 
perhet Radio — Signal Substitution 



CHAPTER 7 

Typical Transistor Receiver 122 

Transistor Broadcast Receiver — Operation When No Signal is Be- 
ing Received — Operation When a Signal is Being Received — Volt- 
age and Resistance Checks — Signal Substitution Tests 



CHAPTER 8 

Tuned Radio-Frequency Receiver 149 

Typical TRF Receiver 



Chapter 1 

HETERODYNING ACTIONS 



The mixer circuit combines two different radio-frequency cur- 
rents in order to obtain a third current, usually of lower fre- 
quency, whose modulations will be a true representation of the 
modulations carried by one of the two original currents. In re- 
ceivers one of these two original currents is called the carrier, 
or signal, current. This current is the one which is being received 
from a transmitting station. The other of the two original cur- 
rents is generated within the receiver by a separate oscillator 
circuit and is usually referred to as the local oscillator current, 
or oscillator current. 

In general, it is less difficult to construct circuits that will 
amplify low-frequency currents than it is to construct them for 
high-frequency currents. Thus, in receiver work it is almost 
universal practice to "step down" the carrier frequency by some 
process such as mixing two currents and getting a third current, 
before amplifying the signal. 

TRIODE MIXER 

The triode mixer circuit, while not too common in present-day 
receivers, is the basic mixer circuit. Figs. 1-1, 1-2, and 1-3 demon- 
strate the operation of this circuit. 

Identification of Components 

The components which make up this circuit are identified as 
follows: 



CI — Grid tuned-tank capacitor. 

C2 — RF filter capacitor. 

C3 — Plate tuned-tank capacitor. 

Tl— First RF transformer. 

T2 — Second RF transformer. 

T3 — IF transformer. 

VI — Triode vacuum tube used as mixer. 

Ml — Antenna. 

M2 — Grid biasing-voltage source. 

Identification of Currents 

Four different electron currents, each flowing at a different 
frequency are present in Figs. 1-1, 1-2, and 1-3. These electron 
currents are each shown in color and are identified as follows: 

1. Current at carrier or signal frequency (red). 

2. Current at oscillator frequency (blue) . 

3. Current at difference frequency (solid green) . 

4. Current at sum frequency (dotted green) . 

Circuit Operation 

The first current to be discussed is the radio-frequency carrier 
current being received at the antenna. This current carries the 
modulation, or intelligence, which is intended to be amplified 
and demodulated. Although the primary of Tl has been shown 
as connected directly to the antenna, the signal current may be 
passed through an RF amplifier stage before being subjected to 
the mixing or frequency-changing process. (The mixer stage 
was formerly referred to as the "first detector," to differentiate 
it from the demodulator stage; however, this terminology is 
seldom used in modern practice.) 

Fig. 1-1 shows the currents which will flow in this circuit when 
the local oscillator is not energized. (The local oscillator is con- 
nected to the primary winding of T2.) Only a single half -cycle 
of operation is shown in Fig. 1-1; it is the half-cycle when the 
signal current makes the control grid negative, thereby cutting 
off any current flow through the tube. However, the previous 
half-cycle would have permitted a pulsation of current to flow 
through the tube, and consequently, the remnants of this tube 
current are seen flowing off filter capacitor C2 and exiting down- 
ward through the primary winding of T3 to the power supply. At 
the same time a component of filter current at this frequency is 
shown re-entering the lower plate of C2. 

The tuned tank, composed of C3 and the primary of T3 in the 
plate circuit, is not resonant at this carrier frequency; therefore 



no oscillation will be set up in the tank due to a pulsation of cur- 
rent which repeats itself at this frequency. 

Fig. 1-2 shows a half -cycle of circuit operation when there is 
a local oscillator current but no signal current. The local oscillator 
current is shown (in blue) flowing through the lower, or pri- 
mary, winding of T2, thereby supporting a current at the same 
frequency in the grid tank circuit. 

The half-cycle chosen for Fig. 1-2 is one which makes the grid 
positive so that plate current is flowing through the tube. This 
current is shown entering the top plate of filter capacitor C2 
before flowing to the power supply through the primary of T3. 

Fig 1-3 shows the conditions when the carrier and oscillator 
currents both exist simultaneously in the grid circuit. As an 
example of the frequencies involved, if the carrier signal being 
received were from a broadcast station operating at 1,000 kc 
per second, the local oscillator would be operating at a higher 
frequency, such as 1,500 kc per second. Thus, a million complete 
cycles of the current shown in red will flow back and forth in the 
grid tank each second. Also, it is known that one-million-five- 
hundred-thousand cycles of the current shown in blue will flow 
in the grid-tank circuit each second. Each one will attempt to 
act independently in turning the electron stream on and off in 
the tube. When the two currents are momentarily in phase they 
will be aiding each other. Conversely, when they are momentarily 
out of phase with each other, or in "opposite phase," they will 
oppose and neutralize each other. 

Pulsations of current are released through the tube at both 
frequencies, although the sizes of the pulsations will vary from 
cycle to cycle, depending on the extent to which each of the 
grid driving current-voltage combinations is augmented or can- 
celled by partial combinations with the other. The size of filter 
capacitor C2 is chosen so that it will have low reactance at 
both frequencies. Thus pulsations are effectively filtered back 
to ground through this capacitor, while the electrons themselves 
may continue their journey to the power supply after passing 
through the primary winding of T3. 

The formula for determining the reactance of a capacitor at 
any frequency is: 

-i 



2*fC 
where, 

X c is the capacitive reactance in- ohms, 

f is the frequency of the current in question in cycles per second, 

C is the capacitance in farads. 




Fig. 1-1. Operation of the triode mixer circuit— negative half -cycle, 
no oscillator current. 



PLATE 

CURRENT 

AT OSCILLATOR 

FREQUENCY 




POWER 
SUPPLY 



Fig. 1-2. Operation of the triode mixer circuit — positive half -cycle, 
no carrier current. 



10 



CARRIER 
CURRENT 



DIFFERENCE -^ 

FREOUENCY (Vl) 

PI ATF ^-S 



PLATE 

CURRENT 
AT FOUR 
FREQUENCIES 




i — », ^ IRE SONANT 

B ummi Rn TANK 
1 ♦ CURRENT 

% ®k Jilt 



FILTER 
CURRENT 
AT THREE 
FREQUENCIES 



Fig. 1-3. Operation of the triode mixer circuit- 
currents both present. 



POWER 
SUPPLY 



w 



OUTPUT 

CURRENT 



:arrier and oscillator 



The capacitive reactance is a measure of the opposition which a 
capacitor will offer to electron flow. From this formula it can 
be seen that the opposition varies inversely to both the frequency 
and the size of the capacitor. Thus, any capacitor offers less op- 
position to the flow of high-frequency currents than to those of 
lower frequency, and also a larger capacitor offers less opposition 
to the flow of any current at any frequency than does a smaller 
capacitor. 

In addition to pulsations of current going through the tube 
at the carrier and the oscillator frequencies, pulsations also occur 
at two other principal frequencies, called the sum and difference 
frequencies. It can be demonstrated mathematically that when two 
sine waves of differing frequencies are combined, or added, these 
two additional frequencies, along with numerous others of lesser 
importance in this case, will be created. The sum of the two 
original frequencies is 2,500 kc. In Fig. 1-3 a new current is shown 
completing the filtering process through C2, and flowing out to 
the power supply essentially as DC. This current is shown in 
dotted green, and is labeled as the sum frequency. Since its fre- 
quency is considerably higher than either of the original fre- 
quencies, we know from the foregoing reactance formula that the 



11 



pulsations at this frequency will be filtered by C2 with even 
greater ease than are the original frequencies. 

A fourth current has been shown in solid green in Fig. 1-3. 
This current is intended to represent the difference frequency. 
Using the 1,000-kc carrier frequency and the 1,500-kc oscillator 
frequency, as before, the difference frequency is 500 kc. Capacitor 
C3 and the primary of T3 are tuned to be resonant at this fre- 
quency; hence, the recurring pulsations will build up a sizable 
tank current at this frequency. Transformer action across T3 will 
build up a current-voltage combination at this same difference 
frequency in the grid circuit of the next amplifier stage. 

All four currents are shown crossing the tube in Fig. 1-3. This 
signifies that the plate current carries pulsations at all four of 
the frequencies under consideration. The pulsations at the three 
highest frequencies are shown entering the upper plate of filter 
capacitor C2; filtering currents at these frequencies are shown 
moving between the lower plate and ground. However, the fourth 
current (in solid green) , representing the important difference 
frequency, is shown flowing directly to the tuned plate circuit, 
where it arrives in the proper phase to reinforce the oscillation 
which exists there at that frequency. The plate tank is con- 
structed and tuned to be resonant only at this difference fre- 
quency. 

It is a characteristic of mixing or conversion circuits of this 
type that the modulation of the carrier signal, which represents 
the desired intelligence being received is conveyed or transplanted 
intact to the new lower -frequency current, called the difference- 
frequency. Thus, in later stages of amplification, even though a 
lower frequency is being amplified, the intelligence of modula- 
tion it carries is a faithful reproduction of the original modulation 
which will eventually be demodulated from the difference fre- 
quency. The output current shown in Fig. 1-3 is caused to flow 
by the transformer action of T3. It flows at the same frequency 
as the difference frequency and has the same modulation char- 
acteristic, or "envelope," as existed on the original signal current 
received from the station. 

The output current-voltage combination from a mixer circuit 
is usually referred to as the intermediate frequency, or IF, be- 
cause it is between the signal current and the audio range in 
the frequency spectrum. 

Unlike most of the circuit diagrams in this text, Fig. 1-3 de- 
picts many cycles of the three higher frequencies, while showing 
a single half -cycle of the lowest, or difference frequency, in the 
plate circuit resonant tank. It is for this reason that arrows are 
shown in both directions in the grid and filter circuits. 

12 



PENTAGRID MIXER CIRCUIT 

Another effective means of mixing voltages of two different 
radio frequencies is shown in Figs. 1-4, 1-5, and 1-6. This arrange- 
ment involves the use of two vacuum tubes. 

Identification of Components 

The pentagrid mixer circuit contains the following components: 

Rl — Grid-drive and grid-return resistor. 

R2 — Cathode-biasing resistor. 

R3 — Screen-grid dropping resistor. 

R5 — Power-supply decoupling resistor. 

CI — RF tank capacitor, working in conjunction with L2. 

C2 — Coupling capacitor between local oscillator and mixer. 

C3 — Screen-grid filter capacitor. 

C4 — Plate-tank capacitor. 

C7 — Power-supply decoupling capacitor. 

LI — Primary winding of input RF transformer. 

L2 — Secondary winding of same transformer. 

L3 — Plate-tank inductor. 

VI — Pentagrid mixer tube. 

The local oscillator circuit contains the following components: 

R4 — Grid-leak biasing resistor. 

C5 — Oscillator tank capacitor. 

C6 — Grid-leak capacitor. 

L4 — Oscillator tickler coil. 

L5 — Oscillator tank coil. 

V2 — Triode oscillator tube. 

Identification of Currents 

There are a total of 11 significant currents during normal 
operation of the circuit in Figs. 1-4, 1-5, and 1-6. These currents, 
and the colors they are shown in are: 

1. Input or "signal" radio-frequency current (solid blue) . 

2. VI plate current (solid red) . 

3. Plate-tank current (dotted blue) . 

4. VI screen-grid current (also in solid red) . 

5. VI screen-grid filter current (also in solid red) . 

6. Oscillator-circuit grid-tank current (solid green) . 

7. Grid-driving current for both tubes (also in solid green) . 

8. Grid-leakage current from oscillator tube (also in dotted 
red). 

9. V2 plate current (also in solid red) . 

13 



10. Feedback current (dotted green) . 

11. High-frequency decoupling currents (in both solid blue and 
solid green). 

Details of Operation 

As is the case with all frequency-mixing, or heterodyning, cir- 
cuits, one of the two voltages to be mixed is the signal voltage; 
and the other one is a voltage generated locally in the oscillator 
portion of the circuit. The purpose of the mixing process is to ob- 
tain a new and lower frequency, because lower frequencies are 
easier to handle and amplify. 



HIGH-FREQUENCY 
PULSATIONS BEING 
FILTERED BY PLATE 
PLATE TANK CAPACITOR 

CURRENT 



|WP|JT GRID TANK 

(CARmER) 
CURRENT 




t 

POWER 

SUPPLY 
DECOUPLING 
CURRENT 



Fig. 1-4. Operation of the pentagrid mixer circuit — local oscillator 

not operating. 



14 



Local Oscillator Not Operating — Fig. 1-4 shows the relatively 
few currents which flow when the local oscillator is not operating 
and a signal current is being received. The signal current is re- 
ceived from the antenna and flows in inductor LI. In Fig. 1-4 it 
is shown flowing upward through LI and inducing a companion 
current to flow downward in L2. Since the tank circuit composed 
of L2 and CI is tuned to the particular frequency being received, 
the current induced in L2 will quickly set up an oscillation of 
electrons in the tuned circuit. 

Fig. 1-4 depicts the half-cycle of oscillation when the upper 
plate of CI is made positive. Since the first control grid of VI 
is connected directly to the top of this tank, the voltage on this 

FILTERING 
ACTION OF PLATE 
TANK CAPACITOR 




POWER 
SUPPLY 

DECOUPLING 
CURRENT 



Fig. 1-5. Operation of the pentagrid mixer circuit — no input carrier signal. 

15 



grid will be identical with the voltage at the top of the tank at 
all times. During positive half-cycles, such as in Fig. 1-4, plate 
current through mixer tube VI will be increased. During the 
next succeeding half-cycle, when the top of the tuned tank ex- 
hibits a negative voltage, this same plate current will be reduced. 

Thus the voltage variations of the tuned tank, which are oc- 
curring at the frequency of the carrier signal being received, will 
impose pulsations on the plate current stream at this same fre- 
quency. The plate current flows through plate-tank inductor L3 
and on through the power supply and back to ground. Inductor 
L3 and capacitor C4 form a resonant tank circuit at the difference 
frequency, or IF. Consequently, the pulsations in plate current 
which occur at the signal frequency are unable to excite the tank 
circuit into oscillation. 

Since the screen grid which surrounds the second control grid 
within the tube is connected through a resistor to the positive 
power supply, it will attract and "capture" a large number of 
electrons from the plate-current stream going through the tube. 
These captured electrons become the screen-grid current {also 
shown in solid red in Fig. 1-4) which flows through R3 and re- 
joins the plate current as it enters the power supply. 

No Signal Conditions — Fig. 1-5 shows the currents which flow 
when the local oscillator is operating, but no carrier signal is be- 
ing received. The oscillator is a standard type, known as a tickler- 
coil oscillator, in which energy is fed back from the plate to the 
grid circuit in the appropriate phase to support the oscillation. 

The oscillator works in the following manner. When positive 
voltage is first applied to the plate of V2, it draws an initial surge 
of plate-current electrons across the tube. This current must 
flow through L4 on its way to the power supply. As if surges, or 
accelerates, upward through L4, it induces, by the electromagnetic 
induction process which occurs between inductors, a companion 
current to surge downward through L5. This is the tickler ac- 
tion, and the induced, or companion, current becomes the feed- 
back current which is shown in dotted green. This feedback 
current moves in the downward direction and tends to remove 
electrons from the upper plate of tank capacitor C5 and deliver 
them to the lower plate. The preceeding makes the upper plate 
positive; the control grid of V2 is made positive at the same time 
by the action of the grid driving current which is attracted upward 
through grid resistor R4 by the positive tank voltage. 

The initial surge of feedback current in L5 sets up the con- 
dition of resonance in the tank circuit, which is tuned to a new 
frequency that is known as the local oscillator frequency. Thus, 
on the next succeeding half -cycle, the top of the tank will be at 

16 



a negative voltage, and the control grid will stop the flow of plate 
current through V2. The tube thus conducts intermittently rather 
than continuously, and the plate current is a special case of 
pulsating DC. 

The feedback current generates and supports the oscillating 
tank current (shown in solid green) . In addition to driving the 
control grid of V2, this tank current and its companion tank 
voltage also drive the second control grid of mixer tube VI. When 
the voltage on the top of C5 is positive (Fig. 1-5), it attracts 
electrons toward it from both directions, namely, upward through 
resistors Rl and R4. Thus, the control grids to which these re- 
sistors are attached will be made positive. On a negative half- 
cycle, when the voltage at the top of C5 is negative, both grid 
driving currents will be driven away from the tank, and conse- 
quently, will flow downward through the two resistors, making 
both control grids negative. 

The action of the grid driving current in flowing up and down 
through Rl affords a means of imposing pulsations on the plate - 
current stream through tube VI; these pulsations will occur at 
the local oscillator frequency. Since the plate tank is tuned to a 
much lower frequency, these pulsations of plate current do not 
excite the plate tank into oscillation. All of the plate current 
flows through the inductor and on to the power supply, being 
joined at the entrance to the power supply by the screen-grid 
current previously described. 

The pulsations in the plate current are filtered by plate-tank 
capacitor C4. This applies to both of the special cases just de- 
scribed. The local oscillator and the carrier signal frequencies are 
considerably higher than the resonant frequency of the plate-tank 
circuit. Resonance is defined as a condition wherein the capacitive 
reactance is equal to the inductive reactance. As the frequency 
is increased, the reactance of the capacitor decreases and the re- 
actance of the inductor increases. This is stated by the two re- 
actance formulas, which tell us that capacitive reactance is in- 
versely proportional to frequency and that inductive reactance 
is directly proportional to frequency. 

Figs. 1-4 and 1-5 show an individual pulsation of plate current 
flowing momentarily onto the upper plate of tank capacitor C4 
and driving an equal number of electrons out of the lower plate 
toward the power supply. During the time period between two 
successive positive half-cycles, the actual electrons which made 
up the pulsation in the first place will be drawn off the upper 
plate of C4 and into the power supply by flowing through in- 
ductor L3. Also, during this period, the filter action completes 
itself, and electrons flow back into the lower plate of C4. 

17 



The filter current in Fig. 1-4 is shown in blue, since it flows 
at the same frequency as the carrier signal being received. In 
Fig. 1-5 it is shown in green, since it flows at the frequency of 
the local oscillator. 

Some arrangement must be provided for bypassing these pulsa- 
tions around the power supply. The most common arrangement 
for performing this function is the simple resistor-capacitor com- 
bination known as a decoupling network. R5 and C7 form the 
decoupling network in this example. When the unfiltered pulsa- 
tions reach the junction of these two components, the path through 
the resistor appears as a relatively high impedance, whereas C7 



'NPUT ^DTANK 

(CARRIER) 
CURRENT 



DIFFERENCE 

FREQUENCY 

TANK 

CURRENT 




FEEDBACK 
CURRENT 



0SC TANK 

CURRENT 



GRID- 
GRID- — LEAKAGE 
DRIVING CURRENT 
CURRENT 



Fig. 1-6. Normal operation of the pentagrid mixer circuit. 



18 



has been chosen to have almost negligible reactance or impedance 
at all radio frequencies. Consequently, when each pulsation is 
filtered through the plate tank by C4, it is again filtered past 
the power supply by C7 (Fig. 1-4 and 1-5) . 

The triode oscillator tube operates under Class-C conditions, 
which means that the tube conducts less than 50% of each cycle. 
This is accomplished by using grid-leak biasing. The grid-leakage 
current has been shown in dotted red in Fig. 1-6. During each 
positive half-cycle in the oscillator tank, the grid of V2 will be 
made positive, attracting some electrons from the plate-current 
stream going through the tube. These electrons, once they strike 
the control grid, cannot be re-emitted within the tube; they must 
exit from the tube and flow back to ground through grid resistor 
R4. Because of the high value of R4, the electrons cannot flow 
immediately to ground; instead, they will accumulate on the right 
hand plate of C6, building up a permanent biasing voltage until 
they can leak downward through R4. 

The instantaneous grid voltage varies around this value of nega- 
tive biasing voltage. Some electrons flow into C6 during each 
positive half-cycle, and the discharging process through R4 goes 
on continuously. During negative half -cycles of the oscillator 
voltage, the negative grid-driving voltage added to the negative 
grid-biasing voltage will be sufficient to cut off the plate current 
and hold it cut off for more than a half of each cycle. 

Full Operation — Fig. 1-6 shows the sum total of all currents 
which flow in the mixer circuit during normal operation. Basi- 
cally, all the currents which flow in the two separate examples 
discussed previously will also flow during normal operation. The 
plate current through VI will be affected by the varying voltages 
on its two control grids, and the plate current will pulsate simul- 
taneously at each of these two input frequencies. Thus, the two 
filter actions in C4 and C7 will occur side by side with and inde- 
pendently of each other, as shown in Fig. 1-6. 

When two separate voltages of different frequencies are mixed 
in a circuit such as this one, the plate current pulsates at a num- 
ber of frequencies in addition to the two applied frequencies. It 
can be shown mathematically that these pulsations will also occur 
at the sum of the two applied frequencies and at many multiples 
of this sum, such as twice, three times, etc. It can also be shown 
that pulsations will occur at the difference between the two ap- 
plied frequencies. 

All but one of these additional frequencies will be considerably 
higher than the resonant frequency of the plate-tank circuit. The 
diffrence frequency is the lowest one at which pulsations occur 
in the plate-current stream. Hence, if the plate tank is tuned to 

19 



resonate at the difference frequency, all the other frequencies 
will be filtered to ground through C4 and C7, in the same man- 
ner that the two primary frequencies are filtered. 

An oscillation of electrons (shown in dotted blue in Fig. 1-6) 
will be set up in the plate-tank circuit at the difference frequency. 
Each cycle of the oscillation will be reinforced by a single pulsa- 
tion of plate current. The amplifier stages which follow the mixer 
tube are tuned to this same frequency so that it is the only one 
which will be amplified. 

In AM broadcast reception a difference frequency or inter- 
mediate frequency of 455 kc per second is fairly standard. In 
higher-frequency reception, such as FM, television, and radar, 
intermediate frequencies ranging from 1 or 2 mc up to 50 or more 
are not uncommon. 

This feature is important when you consider that most receivers 
are designed to receive any one of several transmitters which are 
radiating over a band of frequencies, rather than on a single fre- 
quency. Thus, the input tank circuit, consisting of L2 and CI, 
has to be tunable; at least one of the components must be variable. 
The normal practice is to use a variable capacitor, as indicated 
in Fig. 1-4, 1-5, and 1-6. 

In the oscillator tuned circuit, we find another variable capaci- 
tor, C5. C5 and CI are usually ganged and varied simultaneously 
by a single tuning control. All circuit elements are designed so 
that no matter what frequency of carrier is being received, the 
oscillator tank-circuit frequency will always differ from it by 
the same amount, such as 455 kc. When this condition is met, 
the plate-tank circuit and all subsequent amplifier-tank circuits 
can be manufactured to resonate at this single frequency without 
the necessity of any tuning adjustments on the operator's part. 
This feature allows the following circuits to be designed for 
maximum gain, sensitivity, etc. 

PENTAGRID CONVERTER CIRCUIT 

Figs. 1-7 and 1-8 show two alternate moments in the operation 
of the pentagrid converter circuit of a radio. This circuit per- 
forms several complex and important functions, as follows: 

1. It receives the incoming radio-frequency signal from the 
antenna. 

2. It generates an entirely separate oscillation frequency. 

3. It mixes or combines these two frequencies into a third fre- 
quency, known as the intermediate frequency (IF). This 
combining process is referred to as frequency conversion, 

20 



4. It amplifies or increases the strength of this new frequency, 
to a level much higher than either of the original frequencies. 

Identification of Components 

The various components which make up this circuit, along with 
their functional titles, are as follows: 

Rl — Grid-leak biasing and driving resistor. 

CI — Automatic volume-control (AVC) storage capacitor. 

C2 — Variable capacitor, controlled by the tuning dial. 

C3 — Trimmer capacitor. 

C4 — Variable capacitor in oscillator circuit, controlled by the 
tuning dial. 

C5 — Fixed oscillator tank capacitor. 

C6 — Coupling and isolating capacitor. 

LI — Radio-frequency transformer, also called the antenna 
transformer. 

L2 — Oscillator inductor, which also serves as an auto-trans- 
former. 

Tl — IF transformer. 

VI — Five-grid vacuum tube used as a frequency converter. 

Identification of Electron Currents 

In order to understand everything that is happening inside this 
circuit you must be able to visualize each electron current at 
work in it. These currents are as follows: 

1. Antenna current (solid blue) . 

2. RF tank current (also in solid blue) . 

3. Driving current for second control grid (dotted blue) . 

4. Oscillator tank current (solid green) . 

5. Oscillator feedback current (dotted green) . 

6. Driving current for first control grid (dotted green). 

7. Tube plate current (solid red) . 

8. Grid-leakage current (dotted red) . 

9. Screen-grid current (also in dotted red) . 

10. IF plate-tank current (also in dotted blue). 

11. Grid-tank current for next stage (also in dotted blue). 

12. AVC current (also in solid green). 

Details of Operation 

Transformer Action — The signal or antenna current in solid 
blue in Fig. 1-7 is caused to flow up and down through the pri- 
mary winding of antenna transformer LI by the so-called radio 
waves transmitted by a radio station. The frequency with which 

21 



it changes direction is, of course, equal to the frequency of the 
signal being received. For example, if your radio is tuned to a 
station that is broadcasting on a frequency of 1,000 kc, the an- 
tenna current makes a million complete journeys up and down 
through the primary winding of LI every second. 

Because of the transformer action between the primary and 
secondary windings of LI, the RF tank current (also shown in 
solid blue) is driven by this antenna current. Inductance is a 
sort of electrical inertia and can be compared with the inertia 
of mechanical devices. For example, it requires extra effort to get 
any large stationary object, such as an automobile, moving. How- 
ever, once it is moving, extra effort is required to bring it to a stop. 

Both of the foregoing effects result from mechanical inertia. 
In electrical inertia, which is called inductance, it is not the mass 
of the electron that concerns us, but the electrical charge carried 
by each electron. The principal characteristic of any inductance 
is that it tries to keep the amount of current flowing through it 
at a constant value. This property leads us to the transformer 
action which occurs between the primary and secondary wind- 
ings of LI. As the amount of antenna current shown in Fig. 1-7 
increases or builds up in the upward direction, the amount of 
RF tank current increases in the downward direction in the 
secondary winding. 



ANTENNA 
CURRENT 



DRIVING CURRENT 

FOR SECOND 

CONTROL GRID 



PLATE 
CURRENT 




Fig. 1-7. Operation of the pentagrid converter circuit — first 
control grid negative, second control grid positive. 



22 



Similarly, in Fig. 1-8, when the antenna current increases in 
the downward direction in the primary winding, the RF tank 
current in the secondary winding increases in the upward direc- 
tion. An inductance or a transformer responds to changes in the 
amount of driving current, rather than to the intrinsic amount 
itself. Since the antenna current is constantly changing direction 
and amount, making a million complete round trips up and down 
through the primary winding each second, it drives the tank 
current down and up through the secondary winding at the same 
frequency. 

The RF tank circuit consists of capacitors C2 and C3 in parallel 
with the secondary winding of transformer LI. In Fig. 1-7 we see 
a condition where a positive voltage, indicated by plus signs, 
exists on the upper plates of these two capacitors. This positive 
voltage exists there because during the previous half cycle, the 
oscillating electrons have all migrated downward through the 
second winding, as indicated by the arrows. 

In Fig. 1-8 a negative voltage exists on the upper plates of these 
capacitors. This negative voltage is indicated by minus signs; 
it results from the fact that during the preceding half-cycle, the 
tank current electrons migrated upward through the secondary 
winding, thus delivering a surplus of electrons at the top of the 
tank. The tank current oscillates continuously between these 




AVC 
VOLTAGE 
(NEGATIVE) 



GRID TANK 

CURRENT 

FOR NEXT 

STAGE 



w osc. 

TANK 
CURRENT 

Fig- 1-8. Operation of the pentagrid connector circuit — first 
control grid positive, second control grid negative. 



23 



capacitors and the secondary winding. It is driven or supported 
in its movements by the antenna current in the primary wind- 
ing. This tank current oscillation results in the top of the tank 
exhibiting a voltage that fluctuates from positive to negative 
at the same frequency as the antenna or signal current. Since 
the top of the tank is connected directly to the second control 
grid of the tube VI, this alternating voltage also exists at that 
control grid and can be used to control, or regulate, the flow of 
plate-current electrons through VI. The control grid acts like a 
throttle valve on the electron stream. When the control grid is 
positive, as in Fig. 1-7, more electrons are allowed to pass through 
the tube. 

When the control grid is negative, as depicted in Fig. 1-8, the 
quantity of plate-current electrons flowing through the tube will 
be reduced accordingly. 

Plate Current — The complete path of plate current, which has 
been shown in red in both Figs. 1-7 and 1-8, starts at the ground 
connection below the three components, L2, C4, and C5. The 
plate current flows through the lower half of inductor L2, then 
out through the center tap and to the cathode of the tube. The 
electrons which make up the plate current are then emitted into 
the vacuum of the tube from the heated cathode. 

The plate current passes between the wires of all five grids 
of the tube and strikes the plate of the tube, which absorbs them. 
They are drawn downward through the primary winding of IF 
transformer Tl, and on to the power supply. The high positive 
voltage of the power supply is the attractive force which draws 
the electrons of the plate current (shown in solid red) along the 
entire path. 

Grid-Leakage Current — Some electrons which are emitted by 
the cathode do not reach the plate of the tube, but strike various 
grid wires and leave the tube as grid current. The grid leakage 
current (shown in dotted red) which flows downward through 
resistor Rl is one such grid current. The complete path of this 
grid-leakage current takes it downward through Rl to ground 
and then upward through the lower half of L2 to the cathode 
of the tube, where it is again emitted into the tube. 

In continuously flowing downward through Rl, this electron 
current developes a voltage across Rl which is known as a grid- 
leak bias voltage. This voltage is more negative at the top of Rl 
than at the bottom. This is confirmed by the fact that electrons 
always tend to flow away from an area of more negative voltage 
toward an area of less negative voltage. This grid-leak voltage 
provides a stable and fixed negative voltage at the first control 
grid. 

24 



Oscillating Tank Voltage — The grid leak bias voltage is not the 
only one on the first grid. There is an oscillation of electrons 
occurring in the tank circuit which consists of inductor L2 in 
parallel with capacitors C4 and C5. The oscillating current has 
been shown in solid green. Since the top of this tank circuit is 
connected to the first control grid by means of C6, the oscillating 
tank voltage is said to be coupled to this grid. In Fig. 1-7 when 
the voltage at the top of this tank circuit is negative, electrons are 
driven upward into the lower plate of C6; this action, in turn, 
drives other electrons away from the upper plate of C6 and 
downward through Rl. The downward flow of electron current 
through Rl tells us that the top of the resistor is more negative 
than the bottom during the particular instant represented by the 
diagram. 

An opposite set of conditions is depicted in Fig. 1-8. The tank 
current (shown in solid green) has reversed itself and flows 
downward through L2 to the bottom plates of C4 and C5, making 
them negative with respect to the upper plates. An electron de- 
ficiency has been created on the upper plates so that they now 
exhibit a positive voltage (indicated by plus signs) . This posi- 
tive voltage draws electrons downward from the lower plate 
of C6, and in turn, draws other electrons upward through biasing 
and driving resistor Rl and downward to the upper plate of C6. 
The upward flow of electron current through Rl tells us that the 
top of this resistor is more positive than the bottom during this 
particular half-cycle of oscillation. 

The current (shown in dotted green) which flows up and down 
through Rl is labeled as the grid driving current because it 
"drives" the first control grid by developing an alternating voltage 
at that grid. This current flows simultaneously with the grid- 
leakage current, but is completely independent of it. Each cur- 
rent develops its own voltage at the grid; the total voltage at this 
grid at any instant of time is the algebraic sum of these two sepa- 
rate voltages. The resulting voltage at the first control grid will 
fluctuate at the same frequency as the oscillator tank current. 
The result will be that the electron stream flowing through VI 
will be increased and decreased at this same frequency. During 
the half-cycle represented by Fig. 1-7 when the control grid is 
made most negative, the electron stream, which is in reality the 
plate current through the tube, will be "throttled back" to a mini- 
mum amount. 

During the half-cycles represented by Fig. 1-8 when the first 
control grid is made least negative, this plate current stream 
will be "turned up" to a maximum value. Hence, the plate current 
through VI fluctuates at the oscillator frequency. 

25 



Sustaining Oscillations — The tank circuit, consisting of L2, C4, 
and C5, along with the cathode and first control grid of the tube, 
make up the most essential parts of a conventional Hartley oscil- 
lator. No oscillation of electrons can continue to exist unless there 
is some form of feedback between the output and the input cir- 
cuits to replenish the inevitable losses. In Figs. 1-7 and 1-8 feed- 
back is obtained by means of the autotransformer action which 
occurs between the lower portion of inductor L2 and the entire 
inductor. 

Oscillator feedback current has been shown in dotted green in 
Figs. 1-7 and 1-8. In Fig. 1-8, it is shown in phase with the oscilla- 
tor tank current (solid green) flowing downward through entire 
inductor L2. Because the two currents are in phase, the feedback 
current reinforces or strengthens the oscillator tank current. 

Autotransformer Action — It is important to understand what 
causes feedback current to flow in the first place. In Fig. 1-8 when 
the top of the tank is positive, the control grid reaches its least 
negative (or most positive) voltage. This permits a surge of plate- 
current electrons to flow from the cathode into the tube. These 
electrons must first be drawn upward from ground and through 
the lower portion of L2. Any inductor will always oppose any in- 
crease or decrease in the amount of current which is flowing 
through it. Consequently, when the plate current flowing upward 
through the lower portion of L2 is increased, a separate electron 
current will be generated in the entire inductor, increasing in 
the downward direction. It is by autotransformer action such as 
this that an inductor tries to keep the total current through it 
from changing. This new current is the feedback current. In 
Fig. 1-7 when the top of the oscillator tank is negative, the volt- 
age at the first control grid has its most negative value. This re- 
stricts the flow of plate current through the tube, and of course, 
it reduces the upward flow of current through the lower portion 
of L2 by the same amount. The inductor responds in the con- 
ventional and expected manner and this time generates a current 
which flows upward through the entire inductor at an increasing 
rate. In this way the inductor succeeds (at least momentarily) 
in keeping the total current from decreasing. This newly gen- 
erated current again acts as the oscillator feedback current; since 
it is flowing in the same direction as the tank current, it rein- 
forces it. Thus, the tank current is strengthened or replenished 
during each half-cycle of operation by autotransformer action. 

The tank circuit is considered to be the input circuit, because 
it drives the control grid. The lower portion of L2 is considered 
part of the output circuit, because plate current flows through it. 
Thus, the lower portion of L2 is a part of both the input and out- 

26 



put circuits. The Hartley oscillator, of which this circuit is an 
example, satisfies the general requirement that there be some 
feedback of energy from the output circuit to the input circuit 
for an oscillation to sustain itself. 

Obtaining the Intermediate Frequency — We have seen that two 
different control grids act as individual throttle valves on the 
plate-current electron stream which passes through the tube. 
The oscillations of electrons in the two tank circuits driving these 
grids are occurring at different frequencies. Both circuits are 
tuned by turning the familiar tuning dial on the front of the 
radio. The circuit components are chosen so that the oscillator 
tank which drives grid number 1 (the lower grid) will always 
oscillate at a frequency which is 455 kc higher (faster) than that 
of the antenna current being received. This antenna current, of 
course, supports the RF tank current which drives the second 
control grid (grid number 3) . 

The plate current through the tube will obviously fluctuate 
at each one of the two basic frequencies; however, it will also 
fluctuate at many other frequencies which depend on these fre- 
quencies. Most important are the sum of the two basic frequencies 
and the difference between them (455 kc) . In addition, it will 
fluctuate at a frequency which is twice their sum, twice their 
difference, three times their sum or difference, etc. 

We are interested in only one of these many new frequencies — 
the difference frequency of 455 kc— since we have already chosen 
this as the intermediate frequency, or IF, of our radio. The ad- 
vantage in using a single or fixed IF is that each of the tuned 
circuits (normally 4) in the IF stage can be tuned to this one 
frequency and thereafter will require little or no attention. 

Since the plate-tank circuit is tuned to the IF frequency, the 
pulsations in plate current which occurs 455,000 times each sec- 
ond will very quickly excite an oscillation in the tank. This 
oscillating plate-tank current has been shown in dotted blue. 
Fig. 1-8 depicts an instantaneous set of conditions when the elec- 
trons of the plate-tank current are moving upward through the 
primary winding of Tl, thereby making the upper plate of the 
tank capacitor negative with respect to the lower plate. At the 
same instant, a pulsation of plate-current electrons is arriving 
from the tube and making this upper capacitor plate still more 
negative with respect to the lower plate. In this manner the 
pulsations of plate current reinforce the oscillation in the tank 
current. This sequence of events repeats itself 455,000 times every 
second when the radio is tuned to a station. 

The plate-tank current supports another oscillation of electrons 
in the grid -tank circuit of the next stage. This is accomplished 

27 



by transformer action between the primary and secondary wind- 
ings of Tl and will be discussed more fully in a later chapter. 

Screen-Grid Current — The wires of grids number 2 and 4 con- 
stitute a screen grid, which "screens" or isolates the second con- 
trol grid from the first control grid and from the plate circuit 
of the tube. The screen grid is connected to the high positive 
voltage of the power supply; therefore it also attracts electrons 
from the area around the cathode. Most of these electrons pass 
through the screen-grid wires and eventually strike the plate. 
However, some of the electrons passing through the tube will ac- 
tually strike the screen-grid wires and will exit from the tube 
as screen-grid current, or screen current. This current has been 
shown in dotted red up to the point where it rejoins the plate 
current. Then both currents are shown in solid red as they pro- 
ceed along the B+ line of the radio to the power supply. 

AVC Current and Voltage — The final current which flows in 
the circuit of Figs. 1-7 and 1-8 is the one associated with the 
function known as "Automatic Volume Control" (AVC). This 
current and the resulting AVC voltage which exists on the left 
plate of CI have been shown in solid green. The AVC voltage 
acts as a permanent biasing voltage on the second control grid 
of VI. The means by which it is obtained will be discussed more 
fully in Chapter 2. 

The AVC voltage varies from a low to a high value, but it is 
always negative. When the strength of the radio signal being 
received is low, or weak, the antenna current will be weak, and 
the AVC voltage stored on the lower plate of capacitor CI will 
be a low negative voltage. 

When the received signal strength is high, or strong, the an- 
tenna current will be strong, and the AVC voltage will be a high 
negative voltage. 

A weak AVC voltage will increase the gain of the converter 
stage. A strong AVC voltage will reduce the gain of this stage. 

BEAT-FREQUENCY OSCILLATOR 

The beat-frequency oscillator, or BFO, is a special example 
of the heterodyning process by which two voltages of different 
frequencies are heterodyned to produce a third voltage of a much 
lower frequency. In the examples previously studied the input 
carrier signal was amplitude-modulated and that the resultant 
difference frequency also carried the same intelligence modula- 
tion on it. 

A beat-frequency oscillator is required when the input carrier 
signal has been keyed or coded with dots and dashes. This is a 

28 



special type of modulation known as interrupted continuous-wave 
(ICW). The signal can be detected in the ordinary sense by a 
simple diode detector, such as V3 in Fig. 1-9. However, after the 
detection process is completed, there would be no audio voltages 
by which the listener could tell that demodulation or detection 
had occurred. In such a case, the detection process will have 
been wasted. 

When an RF or IF signal which has been keyed or interrupted 
to form the dots and dashes of the well-known Morse code or 
similar intelligence reaches a detector circuit, the diode detector 
will conduct electrons at a constant rate during the periods when 
dots and dashes occur. During the periods when these pulses are 
not occurring, it will not conduct. Electrons flowing at a constant 
rate downward through load resistor R4 will generate a constant 
or DC voltage for as long as they flow (during each dot and dash, 
for example) but this flow will not result in an audio voltage in 
the headphones. Instead, a single surge of current will flow into 
the headphones (downward) at the start of each pulse, and 
another surge of current will flow upward through the head- 
phones at the end of each pulse. The beginning and end of each 
dot and dash will cause a single cycle of "noise" to be heard in 
the headphones. 

Identification of Components 

The following circuit components perform the indicated func- 
tions in this simplified BFO circuit: 
Mixer Components 
Rl — Cathode biasing resistor for VI. 
R2 — Screen-grid dropping resistor. 
CI — Input-tank capacitor (variable) . 
C2 — Screen-grid filter capacitor. 
C3 — Plate-tank capacitor. 
LI — Input-tank inductor. 
L2 — Plate-tank and coupling inductor. 
Oscillator Components 
R3 — Grid driving and biasing resistor. 
C4 — Oscillator tank capacitor (variable) . 
C5 — Grid coupling and biasing capacitor. 
C6 — Oscillator coupling capacitor. 
C7 — Plate-filter capacitor. 
L3 — Oscillator tank inductor. 
L4 — Radio-frequency choke. 
V2 — Triode oscillator tube. 
Detector Components 
R4 — Variable resistor (volume control) . 

29 



R5 — Cathode resistor. 

C8 — RF filter capacitor. 

C9 — Additional RF filter capacitor. 

CIO — Audio coupling capacitor. 

L5 — Output inductor. 

V3 — Diode detector. 

Ml — Output meter for indicating zero beat. 

M2 — Headphones for listening to code. 

Identification of Currents 

There are at least 15 different electron currents working in 
this BFO circuit; these include: 



INPUT 

CARRIER 

TANK 

CURRENT 








TANK 
CURRENT 
AT OSC. 
AND SIGNAL 
FREQUENCIES 

DETECTC 
DRIVE 
L21 CURRENT 



DETECTOR 

PLATE 
CURRENT 

/ 




FEEDBACK 
CURRENT 



y^ssdfei 





HEADPHONE 
CURRENT 



OSC 

TANK 

CURRENT 



Fig. 1-9. Normal operation of the beat-frequency oscillator circuit. 



30 



Mixer Currents 

1. Input signal tank current (solid blue) . 

2. Pentode plate current (solid red) . 

3. Pentode screen current (also in solid red) . 

4. Screen-grid filter current (also in solid red) . 

5. Plate-tank current at signal frequency (dotted blue) . 

6. Plate-tank current at oscillator frequency (dotted green) . 
Oscillator Currents 

7. Oscillator tank current (solid green). 

8. Oscillator grid driving current (also in solid green). 

9. Coupling current to plate tank (also in solid green) . 

10. Triode plate current (also in solid red) . 

11. Feedback current in oscillator tank (dotted red) . 

12. Plate-filter current (also in solid red) . 
Detector Currents 

13. Detector drive current (also in dotted red) . 

14. Detector plate current (also in solid red). 

15. RF filter currents (dotted blue and dotted green) . 

Details of Operation 

Any qualitative analysis of the operation of this circuit should 
begin with the carrier signal which is received at or delivered to 
input inductor LI from a preceding amplifier or the receiver an- 
tenna. The signal current (solid blue) oscillates between LI and 
CI. (The oscillation is supported by transformer action between 
LI and a preceding inductor.) This action and the prior inductor 
are not shown in Fig. 1-9. 

The oscillating tank current alternately makes the control 
grid of VI negative and positive. With conditions as depicted 
in Fig. 1-9, the tank current is flowing downward through LI, 
thereby removing electrons from the upper plate of CI, making 
it and the control grid, positive. 

The positive control grid encourages, or increases, the flow 
of plate current through the tube so that a pulsation of plate 
current occurs, one such pulsation occurring for each cycle of 
the oscillation. Each of these pulsations will support a single cycle 
of oscillation in the plate tank circuit, consisting of L2 and C3, 
The oscillation which is generated and supported by these pulsa- 
tions has been shown in dotted blue. 

The values of plate-tank components have been chosen so that 
they will resonate at approximately the frequency of the incom- 
ing carrier signal. Therefore, a natural oscillation will be set up at 
this frequency whenever a carrier signal is being received. The 
plate current, which has been shown in red, continues through 
L2 and enters the positive terminal of the power supply. It is 

31 



joined at the junction of L2 and R2 by the screen-grid current, 
also shown in red, which exits from the tube at the screen grid 
and flows through screen dropping resistor R2. Any pulsations, 
or surges, which may characterize this screen current as it conies 
from the tube, will be filtered or bypassed harmlessly to ground 
through C2. Fig. 1-9 shows a half -cycle of this filter current 
flowing downward from the lower plate of C2, because a pulsa- 
tion of electrons has just come from the screen grid. On the next 
succeeding half-cycle, when no pulsation is occurring, this filter 
current will flow upward onto the lower plate of C2. 

The oscillator circuit which is constructed around triode V2 
is a conventional Hartley circuit. The tank circuit, composed of 
L3 and C4, is tuned by variable capacitor C4. Component values 
are chosen so that they will resonate at a frequency very close 
to the frequency of the incoming carrier signal. In this respect, 
it differs from the circuits previously described, whose purpose 
was to produce a new frequency (IF) which still falls in what is 
known as the radio-frequency range. In the beat-frequency oscil- 
lator the two input frequencies — the carrier and the local oscilla- 
tor — produce a difference frequency which is in the low audio 
range, from a few cycles to a few thousand cycles per second. 
Furthermore, C4 may be varied so that the difference frequency 
may be varied over a small range. 

The oscillator tank current (solid green) moves continuously 
between L3 and C4. Fig. 1-9 depicts a moment when the electrons 
have all moved to the lower plate of the capacitor, thereby mak- 
ing the upper end of the tank positive. This action draws electron 
current (also shown in green) upward through the grid resistor 
R3. This makes the top of the resistor and the control grid of V2 
positive, and a pulsation of plate current (solid red) will be re- 
leased to flow through V2. It originates at ground below L3, 
flows through the lower portion of L3 then through the tube and 
the radio-frequency choke L5 to the power supply. This cur- 
rent flows intermittently rather than continuously. During that 
portion of a cycle when it is increasing in amount, it induces a 
feedback current in L3 which is also increasing, but in the op- 
posite direction. This feedback current (dotted red) increases in 
the downward direction through L3 when the plate current 
through the lower part is increasing in the upward direction. 

Since the tank current is moving downward through L3 at 
this same time, the feedback current and the tank current are 
in the appropriate phase with each other so that the tank cir- 
cuit oscillation will be sustained by the feedback action. 

In addition to driving the control grid of V2, this oscillating 
tank voltage is coupled via C6 to the plate-tank circuit of tube 

32 



VI. A single half -cycle of this coupling action is depicted in 
Fig. 1-9. Since the top of the oscillator tank is positive at this 
instant, electrons are drawn toward this voltage from the lower 
plate of C6. This action draws an equal number of electrons onto 
the upper plate and away from the plate tank of VI. Since the 
plate tank and the oscillator tank are tuned to almost the same 
frequency, this coupling action will excite a second oscillation 
in the plate tank circuit at the oscillator frequency. This oscilla- 
ting current has been indicated in dotted green. 

The coupling action between the oscillator tank and the plate 
tank could be performed by a straight wire instead of with cou- 
pling capacitor C6. The real function of the capacitor is to isolate 
or block the positive voltage of the power supply from having 
direct access to ground through L3 and L2. 

As we stated in the previous example of a mixer circuit, when 
two frequencies are mixed, a number of new frequencies are 
created — namely, the sum and difference, twice the sura and 
twice the difference, etc. All but one of these are unwanted fre- 
quencies, and arrangements must be provided for filtering them 
all to ground. All of these new frequencies will be inductively 
coupled between L2 and L5. 

The only one of these many frequencies which is wanted is 
the difference frequency. All of the other frequencies, including 
the two original frequencies, have an important distinguishing 
characteristic in common, namely, that they are radio-frequencies 
and may therefore be easily separated from the difference fre- 
quency by simple capacitive filtering. C8 and C9, situated on 
either side of output inductor L5, provide this filtering. 

When the plate of V3 is made positive with respect to its cath- 
ode, the diode will conduct electrons. This diode current is shown 
in solid red. It flows upward from ground through R5, through 
the diode, downward through L5 and R4, and finally through 
indicating meter Ml. 

It is desired that this diode plate current flow only in response 
to the low-frequency "beat note" — which is the difference be- 
tween the two input frequencies. Neither of these input fre- 
quencies, nor any of the other resulting frequencies, will cause 
the diode to conduct. C8 and C9 both provide a very low imped- 
ance path for currents at these frequencies. Fig. 1-9 shows the 
two original frequencies being filtered through the capacitors 
to ground. R5 has a low value, but nonetheless any current which 
flows through the diode must also flow through R5, Consequently, 
R5 constitutes a load to anyone of these higher frequencies which 
may attempt to draw current through the diode. Since C8 has 
lower impedance than R5 to these higher frequencies, each of the 

33 



higher-frequencies draw current into C8 during its positive half- 
cycle, and discharge them during its negative half-cycle. 

Capacitor C4 can be varied to adjust the oscillator frequency. 
This varies the difference frequency which the listener hears in 
the headphones. The difference frequency can be reduced to zero 
by matching the oscillator and carrier signals. When this is done, 
the diode conducts no plate current, and milliammeter Ml indi- 
cates no current through it. This condition is known as "zero 
beat," and is the best method for precisely calibrating unknown 
frequencies against a known frequency. When the two frequencies 
drift by even a few cycles, the resulting low beat note can be 
heard by the operator; this enables him to readjust the oscillator 
frequency to attain the condition of zero beat. 

The diode plate current flows only when dots or dashes are 
being received. It is a pulsating DC rather than a pure DC. The 
pulsations occur at the so-called difference frequency. Each indi- 
vidual pulsation drives a surge of electron current onto the left- 
hand plate of capacitor CIO, and this action drives an equal 
amount of electrons downward through the headphone circuit. 
Between each two successive pulsations, the electron current 
flows back out of the left hand plate of CIO and downward 
through R4 and the output meter Ml. This action draws electron 
current upward through the headphone circuit. 

In communications receivers it is the usual practice to beat the 
oscillator frequency against a fixed carrier frequency, rather than 
against a variable frequency. An example is the 450 kc intermedi- 
ate frequency or IF generated by the mixer circuit. 

REVIEW QUESTIONS 

1. What is the main reason that the the fourth one? What happens to 
heterodyning function is used so the fourth one? 

widely? 4. Explain how a resistor-capacitor 

_ _. , . . ,. _ . combination R5 and C7 filters or 

2. Fig. 1-3 indicates plate current "decouples" plate current pulsa- 
pulsations at four frequencies ^ keeping them from enter- 
and filter currents at three fre- ing the ^^j. supp i y . why is 
quencies. What are the names of such decoupling desirable? 
these frequencies? Are the three ,. „,, , . M , . .._ , 
filter currents pulsating DC or 5 - What ■ the m f m difference be- 
true AC? tween the pentagrid mixer and 

pentagrid converter circuits? 

3. What characteristic of capacitor 6. How is positive or regenerative 
C2 permits it to "bypass" three of feedback in Fig. 1-7 developed to 
these frequencies, but not to pass replenish oscillator current? 



34 



Chapter 2 

SIGNAL DEMODULATION 

AND AUTOMATIC 
CONTROL OF VOLUME 

After the IF signal obtained at the mixer output is amplified by 
the IF amplifier, it must be changed to an audio signal which cor- 
responds to the audio that produced the original modulation at 
the transmitter. This process is called demodulation or detection, 
(Actually, detection is a misnomer which stems from the early 
days of radio when circuits were devised to ''detect" or "discover" 
the presence of a signal from a distant station. Today the term 
detector is widely used and can be considered as synonymous 
with demodulator. The term "second detector" is also used when 
referring to the demodulator stage.) 

It is desirable to provide an automatic means of changing the 
amount of amplification of the signal to compensate for varying 
signal strengths because of differing propagation characteristics. 
The AVC voltage and current were discussed in conjunction with 
the mixer circuit of Figs. 1-7 and 1-8. 

The weak audio signal obtained at the demodulator stage re- 
quires additional amplification before it is usable. In modern 
radio receivers all of the foregoing functions — demodulation, auto- 
matic control of volume (AVC), and audio amplification — are 
performed by a single tube. 

DETECTOR, AVC, AND AUDIO AMPLIFIER 

The 12AV6 tube (V3) in Figs. 2-1 and 2-2 is capable of per- 
forming like two separate and distinct tubes. It is a dual diode- 

35 



triode tube; independent electron currents can and do flow from 
the cathode to the diode plates and to the triode plate. In this 
particular circuit the two diode plates are shown connected 
together so that they function as if they were a single plate. 

Identification of Components 

The components which make up Figs. 2-1 and 2-2, and the 
portion of the circuit which they function in are as follows: 

Detector Circuit 

R4 — Detector load resistor. 

R5 — Detector load resistor and volume control. 

C8 — IF filter capacitor. 

C9 — IF filter capacitor. 

T2 — (Secondary) Final IF transformer and tank capacitor. 

V3 — (Diode portion) Detector tube. 

AVC Circuit 

R3— AVC resistor. 

CI — AVC storage capacitor. 

Audio Circuit 

R6 — Grid driving resistor. 

R7 — Triode plate-load resistor. 

CIO — Audio coupling capacitor. 

Cll — IF filter capacitor. 

C12 — Coupling capacitor to next stage. 

V3 — (Triode portion) Audio amplifier tube. 

Identification of Currents 

The following separate and distinct electron currents are at 
work in these three circuits: 

Detector Circuit 

1. Final IF tank current (dotted blue) . 

2. Rectified current which flows through the diode (solid blue) . 

3. IF filter current (also in dotted blue) . 

AVC Circuit 

4. AVC current (solid green). 

Amplifier Circuit 

5. Audio grid driving current (dotted green) . 

6. Triode plate current (solid red) . 

7. IF filter current (dotted blue). 

36 



Detector Operation 

The final IF tank current (dotted blue) oscillates continuously 
between the second winding of T2 and the unnumbered capaci- 
tor in parallel with it. The plate-tank current of the preceding 
IF stage flows up and down through the primary winding of T2 
and provides the necessary energy to sustain the current flow- 
ing through the secondary winding. 

The final tank current alternately makes the top of the tank 
vary between negative and positive voltage values. Fig. 2-1 
depicts an instant when the tank current is flowing upward 
through the secondary winding and delivering electrons to the 
upper plate of the tank capacitor, thus making it negative. Fig. 
2-2 depicts conditions a half-cycle later when the tank current is 
flowing downward through the secondary winding, taking elec- 
trons away from the upper plate of the capacitor and creating a 
deficiency of electrons, or, more simply, a positive voltage at the 
upper plate of the tank capacitor. 

These voltage changes (negative to positive and back to nega- 
tive again) occur 455,000 times each second — the intermediate 
frequency of your radio. On the half-cycle depicted in Fig. 2-2 
the voltage at the top of the tank makes the two diode plates of 
V3 positive; therefore they attract electrons from the cathode 
within the tube. These electrons become the so-called diode cur- 
rent, or rectified current, of the circuit. This current is shown 
in solid blue; it flows through the diode portion of the tube only 
when the diode plates are more positive than the cathode. The 
complete path of these electrons takes them downward through 
the secondary winding of T2, and on through resistors R4 and R5 
to the common ground connection, from where they have a ready 
return access to the cathode of the tube. This diode current is a 
pulsating DC. 

The electrons which make up the diode current are prevented 
from flowing immediately downward through R4 and R5; instead 
they accumulate on the upper plate of C9, building up a small 
negative voltage there. This negative voltage is indicated by 
minus signs in both Figs. 2-1 and 2-2. Even though diode cur- 
rent is not flowing in Fig. 2-1, this negative voltage persists on 
capacitor C9 and continues to discharge its electrons downward 
through the resistors. Consequently, although electron current 
flows intermittently through the diode with 455,000 pulsations 
occurring each second, it flows continuously downward through 
the resistors. 

IF Filter Current — The pulsations in the diode current which 
are occurring at the intermediate frequency are filtered out be- 

37 



TRIODE PLATE 
CURRENT 
(MAXIMUM) 



FINAL DETECTED OR (V3 

IF TANK RECTIFIED CURRENT 

CURRENT DOES NOT FLOW 



IF FILTER 
CURRENT 

a 




AUDIO GRID DRIVING 
CURRENT FOR V3 TRIODE 



Fig. 2-1. Detector, AVC, and audio-amplifier circuit — negative half -cycle 
of IF tank voltage, positive half-cycle of audio voltage. 

tween C9 and the ground. Fig. 2-2 shows a single half-cycle of 
this current being driven into ground, as a pulsation of diode 
current flows onto the upper plate of C9. Fig. 2-1 shows the next 
succeeding half -cycle of filter current flowing back onto the lower 
plate of C9. 

The Audio Voltage — The negative voltage which accumulates 
on the upper plate of C9 is important, since it marks the first ap- 
pearance of the audio voltage — in other words, the intelligence 
or the message, you are trying to receive. This voltage can be 
compared to a pool of negative electrons, with the depth of the 
pool representing the amount of negative voltage. The depth of 
the pool does not change significantly during a single cycle of the 
IF current even though electrons are discharging or "draining" 
continuously downward through R4 and R5 to ground. This is 
because the quantity of electrons coming into or going out of the 



TRIOOE PLATE 
CURRENT 
(MINIMUM) 



FINAL DETECTED OR 

IF TANK RECTIFIED CURRENT 

CURRENT DOES NOT FLOW 

\ 



IF FILTER 
CURRENT 



O 

o 

o 



18: 



i 



+ j£— — INSTANTANEOUS 
_T- TANK VOLTAGE 



AVC 

CURRENT 



4pV^— -> 





_! 



IF FILTER 
CURRENT 



-AUDIO CURRENT 
(PULSATING DC) 



» 



H 



POWER 
SUPPLY 



-IF FILTER 
CURRENT 



AUDIO GRID DRIVING 
CURRENT FOR V3 TRIODE 



Fig. 



2-2. Detector, AVC, and audio -amplifier circuit — positive half -cycle 
of IF tank voltage, negative half-cycle of audio voltage. 



capacitor during a single half-cycle is an insignificant percentage 
of the quantity which is already stored or accumulated there. 

Modulation — When the incoming radio wave carries intelli- 
gence, such as speech or music, it is said to be modulated. The 
term modulation is often used to describe or refer to this in- 
telligence, and the purpose of a detector (also called demodu- 
lator) is to extract this information or modulation from the radio 
signal current. As mentioned previously, the point at which this 
extraction process occurs is at the junction of R4 and C9, and 
the electron pool on the upper plate of C9 marks the first appear- 
ance of the modulation voltage. This voltage varies at frequencies 
which are within the range of the human ear — in other words, 
audio frequencies. 

Fig. 2-3 shows a typical modulated waveform. The modulation 
process which occurs at the transmitter consists of a periodically 



varying or changing of the strength of the individual cycles of 
the signal current so that the radio-frequency signal current can 
be made to " carry" the audio information from the transmitter 
to the receiver. This carrying process has given rise to the almost 
universal practice of calling the modulated radio-frequency sig- 
nal the carrier. 

The final IF tank current which causes the diode portions of 
V3 to conduct electrons varies in strength from one cycle to the 
next in accordance with the modulation pattern. When the tank 
current is strong, as it is during the modulation peaks shown in 
Fig. 2-3, larger pulsations of diode current will flow. These larger 
pulsations deliver more electrons into the pool of electrons ac- 
cumulated on C9. Since there will be several hundred or thousand 
individual pulsations of this rectified current flowing through the 
tube during the period occupied by a single audio modulating 
cycle, the depth of the electron pool on C9 will increase as a mod- 
ulation peak approaches. 

When the individual cycles of the final IF tank current are 
reduced in strength by the approach of a modulation trough, the 
corresponding pulsations of rectified current flowing through the 
diode portion of V3 will also be reduced. This results in fewer 
electrons being delivered into the storage pool on C9. Thus, more 
electrons discharge downward through R4 and R5 and flow in 
during this period, and the negative voltage on C9 decreases as 
the modulation trough approaches. 

This process continues as long as a modulated signal is being 
received, with the result that the negative voltage on C9 (repre- 
sented by the electrons in storage on the upper plate) increases 
with each modulation peak and decreases with each modulation 
trough. In other words, this negative voltage rises and falls at 
an audio rate, and is therefore, by definition, an audio voltage. 
This audio voltage first appears on the upper plate of C9. 

The current which drains continuously downward through 
R4 and R5 from this point is driven downward by the amount 
of this voltage. Consequently, it pulsates at the same audio rate 
or frequency. This is another of the many possible forms of 
pulsating DC which exist in a radio during normal operation. 
It is interesting to note that the electrons which accumulate on 
C9 arrive there as pulsating DC from the diode plates of tube V3. 
These pulsations are occurring at the intermediate frequency, or 
in other words, 455,000 pulsations each second. These same elec- 
trons must eventually leave there, also as pulsating DC, but now 
the pulsations are occurring at whatever audio frequency is being 
received at the moment — a few hundred or at most a few thou- 
sand pulsations each second. 

40 



UNMODULATED 
CARRIER 



MODULATION 
PEAKS 




MANY INDIVIDUAL 
RFORIF CYCLES 

Fig. 2-3. Sine-wave representation of carrier-signal current or voltage. 

Ohm's Law — Since an audio current is flowing downward 
through R4 and R5, an audio voltage must be developed across 
these resistors. This is in accordance with Ohm's law, which 
states that the current through any resistor must always be pro- 
portional to the voltage across it. The formula for this is: 

E = IXR 
where, 

E is the voltage across a resistor in volts, 

I is the current through the resistor in amperes, 

R is the resistance of the resistor in ohms. 

Coulombs Law — There is another convenient and simple for- 
mula known as Coulomb's law which tells us that the quantity 
of electrons stored in a capacitor must always be proportional 
to the voltage across the capacitor. The formula for this is: 

Q = CxE 

where, 

Q is the quantity of electrons in storage in coulombs (one 

coulomb equals about 6 X 10 la electrons) , 
C is the capacitor size in farads, 
E is the voltage between the capacitor plates in volts. 

The voltage across R4 and R5 which you would compute using 
Ohm's law would be identical to the amount of voltage stored on 
the capacitor. This voltage could also be computed using Coul- 
omb's law, if you knew the number of electrons accumulated 
there. At all lower points on R4 and R5 a proportionately smaller 
voltage will exist, depending on the distance of the point from the 
top of R4. Resistor R5 is variable; therefore, almost any propor- 



41 



tional value of this audio voltage can be tapped off and coupled 
to the control grid of the triode portion of V3. R5 serves as the 
manual volume control and is controlled by the knob on the front 
panel of the radio. 

Amplifier Operation 

Coupling Action— During a modulation peak, when the audio 
current is flowing downward through R5 at its maximum rate, 
some portion of this current will be diverted onto the left-hand 
plate of CIO; this action will drive other electrons away from the 
right-hand plate upward to the control grid of the tube and down 
through grid driving resistor R6 to the common ground. This con- 
dition is depicted in dotted green in Fig. 2-2. This would con- 
stitute a negative half-cycle of audio voltage, because the voltage 
developed across R6 by electrons flowing downward through it 
will be negative at the top and positive at the bottom (since elec- 
trons always flow from negative to positive) . 

During a modulation trough, when the audio current is flowing 
downward through R5 at its minimum rate, electrons which were 
previously diverted into the left-hand plate of CIO will now flow 
back out, and downward through R5 to ground (Fig. 2-1) . This 
action will draw the grid driving current upward through R6, and 
into the right-hand plate of CIO. The upward flow of electrons 
through R6 makes the voltage at the top of this resistor positive; 
consequently, a modulation trough would lead to a positive half- 
cycle of operation of the audio amplifier. 

Triode Plate Current — When the grid voltage is made negative 
during a modulation peak, the flow of plate current through the 
triode will be reduced to its minimum value. When the grid 
voltage is made positive during a modulation trough, this plate 
current will be increased to its maximum value. This latter con- 
dition is depicted in Fig. 2-1. 

The triode plate current is another one of the several pulsating 
direct currents which flow in a radio. After it exits at the plate 
of the tube it flows to the left-hand plate of C12, where its pulsa- 
tions are coupled to the control grid of the following stage. The 
plate current then flows downward through R7 and to the power 
supply. It eventually will pass through the power supply and 
be returned to the common ground, from where it has ready re- 
turn access to the cathode of V3. 

It is worth remembering that all tube currents must inevitably 
be provided with a return path to the cathodes of their respective 
tubes. Normally, the radio chassis provides these return paths. 

IF Filter Currents — There are two additional points in Figs. 
2-1 and 2-2 where filtering action occurs. Capacitors C8 and 

42 



Cll act as filter capacitors for any IF pulsations which were 
not filtered to ground by C9. We have learned that the electron 
current which flows downward through R4 and R5 is flowing in 
pulsations which occur at the audio frequency being demodulated. 
However, there will inevitably be some small pulsations occurring 
at the intermediate frequency of 455 kc. These pulsations will 
divide between C8 and R5 in inverse proportion to the impedances 
offered by these two components. Impedance may be denned as 
the opposition to electron flow which a component offers. 

Since C8 is deliberately selected to be large enough to offer 
very low impedance to a current pulsating at the intermediate 
frequency, most of these pulsations (shown in dotted blue) will 
be diverted onto the top plate of C8, and each such pulsation will 
drive an equal number of electrons from the lower plate of C8 
to ground. During the periods between each two successive pulsa- 
tions, electrons will flow off the upper plate of C8 and downward 
(through R5) to ground, and an equal number will flow up 
from ground onto the lower plate of C8. 

For purposes of analogy, you might assume that 90% of the 
strength of each IF pulsation which reaches the junction of C9, 
and R4 will be filtered to ground through C9, and the remaining 
10% will enter R4. Also, assume that 90% of the remaining 
strength of each IF pulsation which reaches the junction of capaci- 
tor C8 and resistor R5 will be "filtered" to ground through C8, 
and the remaining 10% will enter R5. 

From the foregoing approximations it can be concluded that 
99% of the strength of the original IF pulsations is filtered out 
by C9 and C8, and only 1% flows through R5. This 1%, how- 
ever, is coupled to the control grid of the triode section of V3 
through coupling capacitor CIO. Consequently, the plate current 
stream flowing through this triode, in addition to pulsating at 
the audio frequency, will also exhibit extremely small pulsations 
at the basic intermediate frequency. Thus, 90% of the strength 
of these individual pulsations will be filtered to ground through 
capacitor Cll. 

AVC Operation 

In the discussion of the detector circuit operation it was pointed 
out that the negative voltage pool on the upper plate of C9 
would rise and fall at an audio rate and that it represented the 
first appearance of the audio voltage in a receiver. Since this volt- 
age is never positive, but always negative, it continuously drives 
electrons downward through R4 and R5. Now examine what hap- 
pens along the AVC line, which connects to the junction of 
C9 and R4. 

43 



The most important thing you should understand about the 
AVC line is that it does not lead to ground, except through R4 
and R5, Therefore none of the diode rectifier current, which 
is shown in solid blue, can flow to ground through the AVC 
line. 

AVC Current Flow — Since the AVC line does not lead to 
ground, any current which flows through R3 has to be a two-way 
current — a true alternating current. During modulation peaks, 
when the electron pool on C9 is large, electrons will be driven 
to the left from C9 into the AVC line, and through large-value 
AVC resistor R3. This condition has been depicted in Fig. 2-2. 
It is this current which delivers electrons to the upper plate of 
CI and builds up the stored charge which becomes known as 
the AVC voltage. 

During modulation troughs, when the electron pool on C9 is 
small, the AVC current will flow to the right through the AVC 
resistor. This action tends to discharge the electrons stored on CI. 

From the foregoing, we see that the AVC current is an alter- 
nating current which flows through R3 at the audio frequency 
being demodulated or detected. 

AVC Voltage — Before a radio is turned on, of course, there 
is no AVC voltage stored on the upper plate of CI. However, 
once the detection process begins, electrons will begin accumu- 
lating on the upper plate of CI. This accumulation process can 
result from only one sequence of events, namely, that the AVC 
current which flows to the left through the AVC resistor R3 dur- 
ing modulation peaks (Fig. 2-2) , is larger than the current which 
flows through the resistor to the right during the modulation 
troughs depicted in Fig. 2-1. 

After several hundred audio cycles have occurred, the AVC 
voltage on CI will assume the average value of the peak and 
trough voltages which occur on the upper plate of C9. When this 
happens, the imbalance between the two half-cycles of AVC 
current will disappear. In other words, the incoming AVC cur- 
rent of Fig. 2-2 will bring just as many electrons into the storage 
pool on AVC capacitor CI as the outgoing current of Fig. 2-1 
takes away. 

Resistor R3 is a very large resistor — usually 2.2 or 3.3 meg- 
ohms. CI is also large — approximately .05 mfd. Therefore the 
product of these two values (R X C) will give a very long time 
constant. Using 3.3 megohms, and .05 mfd, the time constant is 
.165 second; thus, it will require perhaps five or ten times this 
long to discharge the AVC voltage, once it has accumulated on 
the upper plate of CI. Obviously, since a single audio cycle com- 
pletes itself in a small fraction of a second, the AVC voltage can- 

44 



not discharge itself during a single modulation trough, or even 
several hundred of them. 

Another way of saying this is that the quantity of electrons 
stored on CI is so great in comparison to the number which flow 
in or out during a single half of an audio cycle, that the amount 
of the AVC voltage cannot be changed by a single modulation 
peak or trough. Still another way of saying this is that the AVC 
voltage does not "respond" to the audio modulation. 

Signal Fading — The AVC circuit is provided to protect the 
listener against undesirable changes in volume as a result of 
changes in signal strength due to propagation anomalies. 

During a signal fade the incoming carrier signal received at 
the antenna will be significantly reduced in strength, This causes 
a proportional reduction in all subsequent derivatives of this 
same current. Therefore, the final IF tank current shown in 
dotted blue in Figs. 2-1 and 2-2 will be reduced. When this hap- 
pens, fewer electrons will be drawn across the diode portion of 
V3 each positive half-cycle, and the audio voltage which appears 
on the upper plate of C9 will fluctuate between lower values of 
negative voltage. In other words, both the peak and the trough 
voltages will be reduced in size when a signal fade — which may 
persist for several minutes — occurs. 

During a fade, a new imbalance is created between the AVC 
current which comes into CI during the modulation peaks and 
which flows away from CI during the modulation troughs. More 
electron current will flow away from CI during a modulation 
trough than flows into it during a modulation peak. After many 
cycles have elapsed, this imbalance partially discharges the 
negative AVC voltage stored on CI, until it eventually again 
assumes the average value of the peak and trough voltages which 
are occurring on the upper plate of C9. When it reaches this new 
average value, the imbalance between incoming and outgoing 
AVC current again disappears. 

A signal fade would ordinarily reduce the audio output of the 
receiver and require adjustment of the volume control in order 
to maintain a comfortable sound level. The negative AVC voltage 
on the upper plate of CI does this job automatically. Since it is 
connected to the control grids of the preceding tubes, it acts as a 
biasing voltage for these tubes. A less negative AVC voltage, 
which results from a signal fade, will increase the gain (ampli- 
fication) of these tubes and will thereby largely nullify the effect 
of the signal fade. 

Signal Build-Up — A signal build-up, which is also caused by 
propagation anomalies, will increase the strength of the antenna 
current; this will cause proportionate increases in the strength 

45 



of each of the subsequent derivatives of that current, including 
the final IF tank current. Each individual pulsation of detector 
current through the diode portion of V3 will therefore increase, 
and the quantities of electrons stored on C9 during each modu- 
lation peak and trough will increase. When this occurs, a new 
imbalance is created between the two alternate half-cycles of 
AVC current. More current will be driven to the left along the 
AVC line during modulation peaks than will flow to the right 
during modulation troughs. This has the effect, after hundreds 
of audio cycles, of delivering more electrons into storage on 
CI, than are taken away so that it recharges Cl to a more nega- 
tive voltage. This imbalance will persist until the AVC voltage 
again equals the average value of the peak and trough voltages 
appearing on C9. 

A signal build would normally increase the output of the re- 
ceiver, again requiring adjustment of the volume control. With an 
AVC circuit, the more negative AVC voltage biases the preceding 
tubes more negatively, reducing their gain and partially nulli- 
fying the original effect of the signal build. 

As long as a radio is turned on and tuned to a station, there 
will be some electrons in storage on the AVC capacitor; conse- 
quently, the control grids of the converter and IF amplifier tubes 
will always have some negative biasing voltage applied to them. 
The amount of this voltage will vary directly with the strength 
of the incoming signal; therefore the gain of the stages preceding 
the detector will vary inversely with the strength of the incoming 
signal. The result is that once the audio output level has been 
chosen by the volume control, the output level will be main- 
tained or adhered to very closely despite fairly wide variations 
in the incoming signal strength. Automatic volume control is 
sometimes called automatic gain control. 

When a radio is tuned so that it is not receiving any station, 
or when it is turned off completely, the electrons stored on C9 
will discharge to ground through R4 and R5 in a fraction of a 
second. The electrons stored on Cl will also discharge to ground, 
but they must first flow through the large resistor R3, and then 
through resistors R4 and R5. This discharging process will there- 
fore require several seconds for completion. 

GENERATION OF POSITIVE AVC VOLTAGES 

One of the many significant differences between vacuum-tube 
receivers and transistor receivers is in the generation and applica- 
tion of voltages for the automatic control of receiver volume 
(AVC), or gain (AGC). 

46 



With vacuum-tube amplifier stages, gain is most readily varied 
and controlled by controlling the grid-cathode voltage relation- 
ship. A decrease in the negative grid-to-cathode voltage will in- 
crease the gain of a vacuum-tube amplifier. A conventional diode 
rectifier may generate either a negative or a positive voltage, 
depending on whether the output is taken from the plate or the 
cathode. In either case, this voltage will vary proportionately with 
the variations in carrier signal strength. 

A positive control voltage, which becomes more positive as car- 
rier signal strength increases, cannot be used for automatic gain 
control in a vacuum-tube stage, because a more positive voltage 
applied to a control grid will increase rather than decrease the 
gain of the stage. However, a negative control voltage which be- 
comes more negative as signal strength increases can be used 
for AVC, since it decreases the gain of a tube amplifier stage. 
This is the essence of the AVC action in vacuum-tube receivers. 

The foregoing restriction does not apply to PNP transistor 
amplifier stages. Because of the directions of electron flow through 
a PNP transistor, a positive control voltage, which becomes more 
positive as carrier strength increases, can be used for AVC pur- 
poses. An increase in the positive biasing voltage applied to the 
base of a PNP transistor will decrease the current through the 
transistor, thereby decreasing the gain of the stage. 

Identification of Components 

In the circuit of Figs. 2-4, 2-5, and 2-6, the following components 
perform the indicated functions: 

Rl — Emitter biasing resistor. 
R2 — Base biasing resistor. 
R3— First IF filter resistor. 
R4 — Variable resistor used as volume control. 
R5— AVC resistor. 
CI — Collector tank capacitor. 
C2 — First IF filter capacitor. 
C3 — Second IF filter capacitor. 
C4— AVC capacitor. 

LI — Primary winding of first IF transformer. 
L2 — Secondary winding of the same transformer. 
L3 — Collector tank inductor (primary of IF output trans- 
former) . 
L4 — Secondary of IF output transformer. 
XI— PNP transistor. 
Ml — Detector diode. 
M2 — Battery power supply. 

47 



Identification of Currents 

The electron currents listed in the following flow during nor- 
mal operation of this circuit. The reader is again reminded that 
his best chance for understanding how a particular circuit oper- 
ates is to be able to visualize the various electron currents in 
motion, and to relate the movements of each current to its 
neighbor. The student can properly claim that he understands 
how a circuit operates only when each current has (1) been 
properly identified, (2) its complete path through the circuit 
traced out, (3) the action which makes it flow understood, and 
(4) the job it does visualized. 

The following 10 electron currents are at work in the circuit 
of Figs. 2-4, 2-5, and 2-6: 

1. Four IF currents (all in solid blue) . 

2. Base-emitter current (solid green) . 

3. Collector -emitter current (solid red) . 

4. Diode current (dotted red). 

5. AVC current (also in dotted red) . 

6. Voltage-divider current (also in dotted red) . 

Details of Operation 

In a transistor, electron flow between the collector and emitter 
is always against the direction the arrow of the symbol points. 
Likewise, electron current from base to emitter must also flow 
against the direction of the emitter arrow. Figs. 2-4, 2-5, and 
2-6 all show these two transistor currents flowing against the 
emitter arrow. 

Two important conditions must be fulfilled for a PNP tran- 
sistor to conduct these electron currents. These conditions are: 

(1) the base must be negative with respect to the emitter, and 

(2) the collector must be negative with respect to both the base 
and the emitter. 

These conditions are fulfilled in transistor XI by connecting 
both the collector and the base to the negative terminal of biasing 
battery M2. Since the emitter is connected to ground (through 
resistor Rl) , the negative bias voltage will tend to drive electron 
current from the base and from the collector into the emitter 
and through Rl to ground. These currents are appropriately in- 
dicated in Figs. 2-4, 2-5, and 2-6. 

The intrinsic voltage difference which exists between the base 
and the emitter at any instant is the most important factor in 
determining the amounts of these two currents which will flow. 
To understand what causes this voltage difference to exist or to 

48 



vary, consider (initially) the movements of three separate elec- 
tron currents. These currents are: 

1. The voltage-divider current (dotted red) . 

2. The base-emitter current (solid green) . 

3. The collector-emitter current (solid red) . 

When this circuit is at rest, meaning when no currents are 
flowing and when the biasing power supply is disconnected, the 
three terminals of the transistor will be at ground or zero volt- 
age, since they are all connected to ground through various com- 
binations of resistors and inductors. However, when power is 
applied to the transistor, the three previously mentioned cur- 
rents begin to flow; each contributes a substantial change in the 
voltage or voltages which exist at one or more of the electrodes. 
It is absolutely necessary to recognize and understand the voltage 
changes which occur at each terminal. In order to do this, each 
of the currents which are associated with these voltages must be 
visualized. 

The voltage-divider current, shown in dotted red, is the sim- 
plest one of the three to understand, since it does not flow 
through the transistor. This electron current leaves the negative 
terminal of the battery, flows to the left through R2, then down, 
and to the right, and upward through R5 and downward through 
R4 to ground. Each point along this path is progressively less 
negative than any point preceding it. 

The base-emitter current, shown in solid green, is also driven 
by the battery. It flows to the left through R2, upward through L2, 
and into the base of the transistor and out through the emitter, 
then downward through emitter resistor Rl to ground. Like the 
preceding current, each point along this path is progressively 
less negative than any point preceding it. Therefore the emitter 
is more negative than ground because of the current flow through 
Rl, but it is less negative than the base, because there is a small 
amount of resistance between the base and the emitter of the 
transistor. 

The collector -emitter current is also driven by the negative 
power-supply voltage (although it is regulated or controlled by 
the amount of current flowing from base to emitter) . This cur- 
rent, which is commonly called the "collector" current, flows 
from the negative terminal of the battery upward through L3, 
then through the transistor from collector to emitter, and down- 
ward through Rl to ground. In flowing through Rl, this current 
develops an additional component of negative voltage across Rl, 
which alters or modifies the biasing conditions existing between 
the base and the emitter. 



In the absence of a received carrier signal, these three cur- 
rents will very quickly stabilize at "equilibrium" values. Each one 
will exert its own particular effect on the biasing voltages of the 
transistor, and each current will become a pure DC. 

When a carrier signal is being received, all biasing conditions 
and all currents are changed. The carrier signal is shown in 
solid blue, and in Fig. 2-4 it is flowing upward through inductor 
LI. The applied emf which causes this upward flow has been 
indicated by a blue plus sign at the top of LI. The resulting in- 
duced current in L2 is also shown in solid blue, and flows down- 
ward. The "back emf" or "counter-emf" associated with this 
current is indicated by a blue minus (negative) sign at the top 
of L2. It is this induced emf which acts as an additional biasing 
voltage at the base of the transistor. 

In any PNP transistor, such as this one, electrons will flow 
from the base and into the emitter. In order for this to happen 
the base must be more negative than the emitter. The base- 
emitter current, shown in solid green, contributes a small amount 




Ave 

CURRENT 
-*- AVC VOLTAGE 



Fig. 2-4. Transistorized IF amplifier, detector, and AVC circuit- 
negative half-cycle of IF. 



50 



of this voltage difference. A much larger component of this 
voltage difference is contributed by the negative voltage induced 
at the top of L2 during any negative half-cycle, such as that 
depicted in Fig. 2-4. Thus the amount of electron flow between 
base and emitter is greatly increased during a negative half- 
cycle. The amount of collector current which flows is greatly 
influenced and regulated by the amount of base-emitter current 
(usually called emitter current) which flows. Thus, the two tran- 
sistor currents (base-emitter and collector-emitter) are both 
substantially increased during a negative half-cycle. 

During a positive half-cycle of RF or IF, both of the transistor 
currents are substantially reduced. This comes about initially 
when the input carrier current (shown in solid blue) induces a 
positive component of voltage at the top of L2. This positive 
component is added to the negative voltage existing at the bot- 
tom of L2 by virtue of the voltage divider current shown in 
dotted red. The net result is a reduced value in the negative volt- 
age at the base of the transistor. Since the PNP transistor re- 
quires a negative voltage at the base for electrons to flow from 
the base to the emitter, it should be evident that any reduction 



BASE 

DRIVING 

CURRENT 



COLLECTOR- 
EMITTER 
'CURRENT 

I INCREASES 



DETECTOR 
CURRENT 
DIODE F LOWS 
DRIVING 
CURRENT 



AUDIO 

OUTPUT 

CURRENT 



IF TANK 7© (SSS& CASSIA ® T 

CER ENT / V - , . \ , ■ [ __ _^^ >8 ^- WT -|-4 i ; 




AVC VOLTAGE 
MORE NEGATIVE 



Fig. 2-5. Transistorized IF amplifier, detector, and AVC circuit- 
signal strength decreased. 



51 



in the negative base voltage will reduce this flow of base-emitter 
current. 

When the base-emitter current is reduced, collector-emitter 
current is reduced proportionately. Normally, the amount of 
collector-emitter current which flows is between 10 and 50 times 
as great as the amount of base-emitter current. It is this property 
of the transistor which enables it to be used as an amplifier, be- 
cause a relatively small amount of change in the base-emitter 
current will cause a substantial amount of change in the collector- 
emitter current. 

Fig. 2-5 shows a positive half-cycle when a relatively weak car- 
rier signal is being received. Fig. 2-6 shows a positive half-cycle 
when the strong carrier signal is being received. Inspection of 
Figs. 2-5 and 2-6 will reveal very few differences in current direc- 
tions or voltage polarities throughout the entire circuit. However, 
the amounts of most currents and voltages will change signifi- 
cantly as we go from a weak signal to a strong signal, or vice 
versa. It is only by visualizing how these changes occur, and by 
recognizing their significance, that the AVC action can be under- 
stood. 

Referring to Fig. 2-5, the pulsations in the collector current 
as it flows upward through L3 will first shock-excite the tank 
circuit, CI and L3, into oscillation, and then each individual pul- 
sation will replenish or sustain a single cycle of this oscillation 
of electrons in the tank. Of course, the pulsations also occur dur- 
ing the negative half-cycles. In Fig. 2-4 a pulsation of electrons 
is shown (in solid blue) flowing upward in the tank circuit. In 
Figs. 2-5 and 2-6 the tank-circuit electrons are returning down- 
ward through L3. 

A companion current is induced in secondary winding L4 by 
each primary winding. During the negative half-cycles (Fig. 2-4) , 
this secondary current is shown as flowing downward through L4; 
during the positive half -cycles (Figs. 2-5 and 2-6), it is flowing 
upward. The back emf associated with the secondary current 
has polarities which are indicated by the appropriate signs at the 
top of L4. The voltage polarities at the top of L4 determine 
whether or not diode Ml will conduct electrons. When the top of 
L4 is positive, as it is during positive half -cycles of RF (or IF) , 
Ml will conduct from right to left. This is the normal flow direc- 
tion for a solid-state rectifier from cathode to anode. This diode 
current is shown in dotted red. These electrons are drawn initially 
out of the upper plate of filter capacitor C2, and flow through the 
diode as indicated, then downward through L4 to ground. 

When no carrier signal is present, the upper plate of C2 will 
charge to a value of negative voltage that is determined by the 

52 



voltage divider current shown in dotted red. This current flows 
continuously from the negative battery terminal through Rl, R5, 
and R4, in that order, to ground. The amount of the initial voltage 
on C2 can be calculated by simple arithmetic using Ohm's law. 

During the weak-signal positive half-cycles of Fig. 2-5, when 
the diode anode is only slightly positive, a small quantity of elec- 
trons are drawn out of this "electron pool" on C2 and flow 
through the diode. This action reduces the negative voltage on 
C2 so that it becomes less negative than the voltage at the right- 
hand end of R3. Consequently, electrons will flow from right 
to left through R3 to equalize this voltage imbalance. Initially, 
these electrons come from the upper plate of C3, also making 
the voltage stored there less negative by a small amount. 

During the strong-signal positive half-cycles of Fig. 2-6, when 
the diode anode is made very positive by the induced voltage at 
the top of L4, a larger quantity of electrons are drawn out of 
the electron pool on C2 and flow through the diode. This action 
substantially reduces the negative voltage on C2 so that a greater 
amount of electrons must flow from right to left through R3 to 
equalize the voltage imbalance. These electrons come from the 
upper plate of filter capacitor C3, reducing the negative voltage 
stored there by a significant amount. 

There are three separate and distinct time periods which must 
be considered when we try to analyze the manner in which an 
AVC circuit operates. These time periods are as follows: 

1. The time required for a single cycle of RF or IF to com- 
plete itself — one or two millionths of a second (microsec- 
onds, in other words) . 

2. The time required for a single cycle of audio-frequency to 
complete itself — a few thousandths of a second (milliseconds, 
in other words) . This time period is several orders of mag- 
nitude longer than one that is measured in microseconds. 

3. The time required for a signal fade or a signal build to occur, 
due to atmospheric or propagation anomalies. This will 
require several seconds or even several minutes to occur. 
This time period is obviously many orders of magnitude 
longer than one that is measured in milliseconds. 

There are three important resistor-capacitor filter combinations 
in this circuit, each one of which is designed to respond to a cur- 
rent/voltage action occurring in a different one of these three 
time periods. These RC filter combinations are as follows: 

1. R3 and C2, which respond only to RF or IF. 

2. R4 and C3, which respond only to audio frequencies. 

53 



COLLECTOR- 
EMITTER 
'CURRENT 

l INCREASES 




AVC 

CURRENT 



Fig. 2-6. Transistorized IF amplifier, detector, and AVC circuit- 
signal strength increased. 

3. R5 and C4, which respond only to sustained changes in re- 
ceived signal strength. 

It is evident in Figs. 2-4, 2-5, and 2-6 that negative voltages 
(meaning electrons) are stored on the upper plates of C2, C3, 
and C4. The voltage level on C2 rises and falls at the intermediate 
frequency to which the final amplifier tank is tuned — 455,000 cps 
is a typical example. On positive half -cycles, electrons are drawn 
out of this capacitor, and an equal number flow up from ground 
and onto the lower plate of C2, On negative half-cycles, electrons 
flow downward onto C2 from R3, recharging C2 and driving an 
equal number of electrons from the lower plate of C2 back into 
ground. The electron current which flows between the lower 
plate of C2 and ground is the principal component of IF filter 
current. (A lesser component which has not been shown will 
flow between the lower plate of C3 and ground.) 

The voltage level on C3 rises and falls at an audio rate. This 
voltage marks the first appearance of the audio voltage in the 
receiver system. Fig. 2-7 shows typical waveforms which relate 
an IF carrier signal to the audio intelligence which it carries. Dur- 



54 



ing modulation troughs, the carrier pulsations are relatively- 
weak, and each pulsation will cause only a small number of elec- 
trons to flow through the diode on the positive half-cycles. Con- 
sequently, the electron pool (negative voltage) on C2 is not de- 
pleted as much, and the electron pool on C3, which replenishes 
the one on C2 will also be depleted only slightly during a modu- 
lation trough. 

The modulation trough is characterized by a succession of 
weak positive half-cycles. The modulation peak, on the other 
hand, is characterized by a succession of strong positive half- 
cycles. Since large numbers of electrons flow through diode Ml 
during each strong positive half-cycle, and since these electrons 



MODULATION 
TROUGHS" 



I. IF WAVE- 
FORMS 



2. PULSATIONS OF 
DETECTOR 
C URREN T 
ZERO VALUE 

OF AUDIO 
VOLTAGE MEAS- 
URED ON C3 

3. AUDIO VOLTAGE 



ZERO VALUE 

OF AVC 
VOLTAGE AS 
MEASURED 
ON C4 



4. AVC VOLTAGE 




Fig. 2-7. Voltage waveforms at four significant points under four different 
operating conditions of the circuit in Figs. 2-4, 2-5, and 2-6. 



55 



must eventually be supplied from the electron pool on C3, it 
follows that the negative voltage on C3 is reduced more during 
a succession of strong positive half-cycles than it is during a 
succession of weak positive half-cycles. Therefore, the voltage 
on the upper plate of C3 is less negative during modulation peaks, 
and more negative during modulation troughs. 

The negative voltage stored on C4 does not change with each 
modulation peak and trough. This voltage will remain constant 
as long as the average signal strength being received is constant. 
If this signal strength fades out or builds up due to propagation 
anomalies, the voltage on C4 will change proportionately, and 
affect the transistor biasing conditions. Thus, the amplification 
provided by the transistor is varied. Consider how this voltage 
change on C4 can be brought about. 

First, consider a signal fade. The waveforms of column 3 in 
Fig. 2-7 indicate that a signal fade is characterized by a long suc- 
cession of weakened IF cycles so that the modulation peaks as 
well as the troughs are reduced in strength. Consequently, dur- 
ing the entire period of a signal fade, fewer electrons must be 
drawn away from C3 to replenish the current which flows 
through Ml. Therefore the average voltage on C3 remains at a 
fairly high negative value. 

The negative voltage on C4, which is the AVC voltage, always 
assumes the average value of the peak and trough voltages which 
are occurring on C3 (after allowing for the steady component 
of voltage developed across R5 by the continuous flow of voltage 
divider current). 

Now consider a signal "build." The waveforms of column 4 in 
Fig. 2-7 indicate that a signal build is characterized by a long 
succession of strengthened IF pulses. The modulation peaks and 
troughs are both stronger, placing greater demand for electrons 
on the negative voltages stored on C2 and C3. The end result 
of a signal build is that the average voltage on C3 will be consider- 
ably less negative than it is during a signal fade. 

During a signal build, the AVC voltage stored on the upper 
plate of capacitor C5, also becomes less negative. Since the base 
of the transistor is connected directly to this point, the amount 
of base-emitter current through the transistor will be reduced, 
and the total amplification which the transistor can provide will 
also be reduced. This decrease in amplification nullifies or com- 
pensates for the adverse effects of the signal build. 

During the signal fade previously discussed, the AVC voltage 
on C4 is more negative than it is during a signal build. This 
serves as a biasing voltage at the base of the transistor and 
increases the amount of base-emitter current which will flow dur- 



56 



ing any positive half-cycles. This increases the amplification which 
the transistor provides, and thereby nullifies or compensates for 
the loss in signal strength due to the signal fade. 

Whenever we consider the response of an RC filter circuit, 
we must inevitably look at the time-constant relationship be- 
tween each filter and the frequency of the current or currents 
under discussion. The time-constant formula states that time 
(in seconds) is equal to the product of the resistance (in ohms) 
and the capacitance (in farads) . The time computed by this 
formula is called the time-constant of the combination. 

It will be instructive to look at the time-constant values for 
these RC filter combinations using component values taken from 
a typical transistor radio. 

These component values are: R3, 510fi; C2, 0.01 mfd; R4, 
5,000n; C3, 0.01 mfd; R5, 39,000fi and C4, 6 mfd. 

The time-constant values of each of these filters is computed 
as follows: 

Tl = R3 X C2 

= 510 x .01 X lO" 6 
— 5.1 X 10 ~ 6 seconds 
= 5.1 microseconds. 

T2 = R4 X C3 

= 5,000 x .01 x lO" 6 
-50 X 10 ~ 6 seconds 
= 50 microseconds or 1/20 millisecond. 

T3 = R5 X C4 

= 39,000 x 6 x 10- 6 

= 234,000 X 10 -"seconds 

= 234 milliseconds. 

The fundamental requirements which are placed on each of 
the RC filters is that each time-constant value shall be long 
when compared to the time for one cycle of one frequency range 
but that it be short when compared to all of the lower frequency 
ranges. By definition, a time constant is considered to be a long 
time constant when its value is five to ten times longer than the 
time duration of one cycle. Under the terms of this definition, 
time Tl is short when compared to a single cycle of IF (455 kc) , 
which requires more than 2 microseconds of time; but it is long 
when compared to an audio cycle. 

Time T2 is short when compared to the time duration of a 
single cycle of audio frequency. The entertainment spectrum of 
audio frequencies ranges from perhaps 50 to 2,000 cycles per 
second. A twentieth of a millisecond is ten times as short as a half 

57 



of a millisecond. Also, time T2 is much longer than the time 
duration of a single cycle of IF (2 microseconds) . It is this long- 
time-constant relationship that enables R4 and C3 to reproduce 
the audio modulation which is carried by the RF and the IF 
signal. 

Time T3, being nearly a quarter of a second in duration, is 
obviously long to even the longest audio cycle. Therefore, the 
AVC filter composed of R5 and C4 does not respond to the modu- 
lation (audio voltage) which appears on the upper plate of C3. 
Time T3 is short however when compared to any of the atmos- 
pheric or propagation anomalies which can cause signal fades or 
builds. These anomalies may last for several seconds or many 
minutes. 

It was stated previously that a PNP transistor amplifier re- 
quires a positive AVC voltage and that an NPN amplifier would 
require a negative AVC voltage. It is evident from line 4 of Fig. 
2-7 that the AVC voltage is negative at all times. This is a seem- 
ing inconsistency, but one which is explainable by our facility 
for using specialized meanings for simple words (and frequently 
without adequate clarification) . In this example, what we mean 
is that a carrier signal must generate an AVC voltage which 
becomes more positive as the signal strength increases, and more 
negative as the signal strength decreases. Columns 3 and 4 in 
Fig. 2-7 clearly show this phenomenon. 

Both the audio voltage generated on C3 and the AVC voltage 
generated on C4 are negative at all times. This is due to the 
action of the voltage-divider current, shown in dotted red, which 
places all points on its flow path at particular values of negative 
voltage, each point being at a less negative value than any up- 
stream point, and at a more negative value than any downstream 
point. Both the audio and AVC voltages must fluctuate around 
their initial negative voltage values. 

On a modulation peak, the audio voltage on C3 must become 
less negative, and during a modulation trough, it must become 
more negative. These relative values are indicated in line 3 of 
Fig. 2-7. 

The AVC voltage on C4 always assumes or "follows" the aver- 
age value of the audio voltage on C3. The student should assure 
himself that he understands how this averaging action takes place. 
This action would be somewhat easier to explain if the voltage- 
divider current could somehow be eliminated from the circuit. 
If R2 were simply removed from the circuit, this voltage-divider 
current could not flow. Then, the audio voltage on C3 would be 
positive instead of negative. Each modulation peak would drain 
more electrons away from C3 to flow through the diode, making 

58 



the C3 voltage more positive. During each modulation trough, 
electrons would be drawn upward from ground through resistor 
R4 and they would reduce this positive voltage somewhat, but 
not completely. 

The AVC voltage on C4 would eventually charge to the average 
value of these peak and trough voltages on C3. This AVC volt- 
age would be a true positive voltage, since it would be the aver- 
age value of two other voltages, both of them positive. During 
each modulation peak, electrons would flow through AVC re- 
sistor R5 from C4 to C3, trying to equalize the two different 
positive voltages. (C3 has the higher positive voltage during the 
modulation peaks.) 

During each modulation trough, electrons would flow from 
C3 to C4, through R5, again trying to equalize the two different 
positive voltages. (C4 has the higher positive voltage during the 
modulation troughs.) 

From this we can see that an electron current which may be 
appropriately labeled as the "AVC current" flows back and forth 
through R5, always flowing from the point of lower positive 
voltage to the point of higher positive voltage. Since this cur- 
rent must flow back and forth with each modulation peak and 
trough voltage occurring on C3, it becomes apparent that the 
AVC current through R5 flows at the audio frequency being 
detected or demodulated. 

Next, it is appropriate to ask why the voltage stored on the 
upper plate of C4 does not change with each inflow and outflow 
of AVC current from or to C3. Why does the AVC voltage on 
C4 not change or vary in accordance with the changes in ampli- 
tude because of the modulation? 

This question can always be answered on a mathematical basis 
by asserting that the RC combination of R5 and C4 is a "long- 
time—constant" to audio frequencies, and will therefore not re- 
spond to modulation. Such a statement is true and unargueable. 
The physical significance of this statement can only be under- 
stood if we consider the quantities of electrons which are in- 
volved in this current movement, and also the quantity of elec- 
trons (or positive ions) which are stored on C4 and which con- 
stitute what we know as the AVC voltage. 

C4 is a 6-mfd capacitor, which means that it is 600 times larger 
than C3 (.01 mfd). When these two capacitors are charged to the 
same value of positive voltage, C4 must, of necessity have 600 
times as many positive ions stored on its upper plate as are 
stored on the upper plate of C3. This is an arithmetical relation- 
ship which is covered by the formula known as Coulomb's law. 
This law states that the voltage across any capacitor is propor- 

59 



tional to the quantity of negative electrons (or positive ions) 
stored in the capacitor. This formula is usually written as: 

Q = CxE 

where, 

Q is the quantity of electrons (or ions) in storage expressed in 

coulombs (one coulomb equals 6 X 10 1S electrons or ions), 
C is the size of the capacitor in farads, 
E is in the voltage across the capacitor in volts. 

From this relationship one can see that if two capacitors of dif- 
ferent sizes are charged to the same voltage, then the larger capa- 
citor will require a larger amount of charge (electrons or ions) 
than the smaller one. 

Again reverting to our hypothetical example, we can see that 
the inflow of electrons from the AVC current to C3 will reduce its 
positive voltage, and the outflow of AVC current electrons will 
increase its positive voltage. The voltage on C4, however, remains 
virtually unchanged when these same quantities of electrons flow 
out and in, respectively. This is because C4 is 600 times larger 
than C3. According to Coulomb's law, a quantity of electrons 
which will change the voltage on CZ by one volt will change the 
voltage on C4 by only l/600th of a volt. The peak value of these 
pulsations will occur during the modulation peaks and the mini- 
mum value during modulation troughs. 



REVIEW QUESTIONS 



1. At what point in Figs. 2-1 and 2-2 
does the "demodulated" audio 
voltage make its first identifiable 
appearance? 

2. In Pig. 2-2, what causes the RF 
filter current to flow away from 
the lower plate of C9? In Fig. 2-1, 
what causes this current to flow 
upward? 

3. During a modulation peak, will 
the instantaneous audio voltage 
on the upper plate of C9 (Fig. 2-1) 
become more negative or less 
negative than during a modula- 
tion trough? 

4. Describe how an audio driving 
voltage is developed across grid 
resistor R6 in Fig. 2-1. 

5. In a typical broadcast receiver 



using AVC or AGC, RC filter cir- 
cuits will be found which exhibit 
"time constant values" making 
them suitable for use in any one 
of three different frequency or 
time domains. Name these three 
time domains. 

What is the frequency of the 
AVC current which is shown in 
solid green in Figs. 2-1 and 2-2? 
Is this a pulsating DC or a true 
alternating current? When this 
receiver is turned off, what hap- 
pens to the electrons which are 
stored in CI and what constitutes 
the AVC voltage? When the re- 
ceiver is first turned on, describe 
the current actions which result 
in the initial build-up of this AVC 
voltage. 



60 



Chapter 3 

NOISE-LIMITING 
PRINCIPLES 



In the transmission of a signal between the station and the re- 
ceiver, noise pulses are often superimposed on the signal. These 
pulses, which may be caused by atmospheric or man-made con- 
ditions, will cause "static" in the output if allowed to pass through 
the receiver. Most amateur and communications receivers employ 
circuits for removing these pulses so that they will not appear in 
the output; however, noise-limiting circuits are seldom employed 
in "entertainment- type" home receivers. In this chapter, three 
types of noise limiters — shunt diode, series diode, and dual diode — 
as well as pentode squelch circuit will be discussed. 

SHUNT-DIODE NOISE LIMITER 

Figs. 3-1 and 3-2 show two separate moments during the opera- 
tion of a simple noise-limiting circuit, which places a diode tube, 
VI, across the grid input circuit of a conventional pentode audio 
amplifier, V2. In the absence of an undesirable noise pulse, the di- 
ode tube does not conduct; this condition might be labeled as the 
7iorvial mode of operation (Figs. 3-1 and 3-2) . When a noise pulse 
is present, the diode tube conducts, as shown in Fig. 3-3, This con- 
duction biases the control grid of the pentode tube beyond cutoff, 
cutting off the tube for the duration of the noise pulse. 

Identification of Components 

The following circuit components in Fig. 3-1, 3-2, and 3-3 per- 
form the functions indicated: 

Rl — Variable resistor (volume control). 

R2 — Variable resistor used as voltage divider to set the noise 
level. 

61 



R3 — Cathode-biasing resistor for V2. 

CI — Input coupling capacitor, 

C2 — Coupling and biasing capacitor. 

C3 — Cathode-bypass capacitor for V2. 

LI — Grid-driving and biasing inductor. 

VI — Diode tube used as noise limiter. 

V2 — Pentode tube used as audio amplifier. 

Identification of Currents 

Three electron currents will flow in this circuit during normal 
operation. Two additional currents are introduced during ab- 
normal operation (when a noise pulse is received) . These cur- 
rents, and the colors used to identify them are: 

Normal Operation. 

1. Input audio current (solid blue). 

2. Voltage divider current (solid green). 

3. Pentode plate current (solid red) . 
Abnormal Operation 

4. Noise current (dotted blue). 

5. Diode current (dotted green). 

Also during abnormal operation, the pentode plate current does 
not flow. 

Details of Operation 

Fig. 3-1 shows a negative half-cycle of operation of the noise- 
limiter circuit when no undesired noise pulse is present. The in- 
put audio signal from the demodulator circuit reaches input 
capacitor CI and drives an electron current downward through 
Rl, producing a negative voltage at the upper end of Rl. Each 
point below the top of Rl will exhibit a lesser negative voltage 
than that at the top during this negative half -cycle, depending on 
the distance. Thus, a movement of the potentiometer arm taps off 
any desired amount of the audio voltage for coupling to V2. 

The physical means by which this coupling action is accom- 
plished is indicated in Fig. 3-1. When current is driven downward 
through Rl, it is also driven onto the left-hand plate of C2. This 
action drives an equal number of electrons out of the right-hand 
plate of C2 and downward through LI. During this downward 
motion of electrons through LI, the top of the inductor will be 
negative in voltage; this is the voltage applied to the grid of V2. 

This pentode will be conducting throughout an entire audio 
cycle, i.e., continuously, during normal operation. This is usually 
referred to as Class-A operation. During a negative half-cycle, 

62 



PLATE 
CURRENT 
(MINIMUM) 




\ DIODE 
BIASING 
CURRENT 



Fig. 3-1. Operation of the shunt-diode noise limiter — negative half -cycle. 

such as is shown in Fig. 3-1, the plate current will be reduced to 
its minimum value. 

Fig. 3-2 shows a positive half -cycle during normal operation of 
this circuit. The input audio current is now being drawn out of the 
left-hand plate of CI; this action draws electron current upward 
through Rl and also out of the left-hand plate of C2 and upward 
through LI. This action places a positive voltage on the control 
grid of V2 and causes the maximum amount of plate current to 
flow through the tube. 

During both the negative and positive half -cycles of operation 
shown in these two illustrations a voltage-divider current (shown 
in solid green) will be flowing continuously through R2 from right 
to left. This current flows continuously in a counterclockwise 
direction through R2 and the power supply in ground, then out 
of ground and back into R2. As the potentiometer arm is moved 
from right to left, a succession of higher and higher positive volt- 
ages will be encountered, and applied to the cathode of VI. This 
potentiometer arm is used to set the noise level at which the cir- 
cuit operates. 



63 



The most important single condition for a diode tube to conduct 
electron current is that the instantaneous voltage at the diode plate 
must be more positive than the instantaneous voltage at the cath- 
ode. The cathode of VI is held at a certain value of positive volt- 
age, depending on the position of the potentiometer arm, so that 
the tube is normally nonconducting. Even the positive half -cycle 
of audio voltage depicted in Fig. 3-2 is assumed to be insufficiently 
strong to make the diode plate more positive than this cathode 
voltage. Consequently, the diode acts as an open circuit during the 
entire audio cycle. The term "open circuit" is frequently used in 
this kind of situation and should be considered as synonomous 
with infinite resistance. Thus, it has no effect on the normal opera- 
tion of the pentode amplifier input circuit. 

Fig. 3-4A shows the RF (or IF) waveform during normal opera- 
tion. This illustration also shows how the same waveform would 
be modified and distorted by a typical noise pulse. Noise pulses 
are characteristically of extremely short time duration and usually 
of very high amplitude. These pulses may be the result of either 
natural or artificial interference — lightning bursts, automobile- 
ignition systems, X-ray equipment, or any one of dozens of other 
types of industrial electronic equipment. 

When the positive half-cycle of even a single cycle of such a 
noise pulse reaches the plate of VI, it makes this plate more posi- 
tive than the cathode, and the diode conducts electrons strongly 
from cathode to plate. The "noise current" which causes this 
condition is shown being drawn upward through LI (in dotted 
blue) in Fig. 3-3 (the resulting diode current is shown in dotted 
green) . This diode current is drawn out of ground below the right- 
hand end of R2 and flows in a short burst (or a series of short 
bursts, depending on the number of cycles of noise voltage pres- 
ent) through the diode from cathode to plate. From the diode 
plate, the current flows onto the right-hand plate of C2, where it 
accumulates and very quickly forms a "pool" of negative voltage 
which biases both the diode plate and the pentode control grid 
negatively. This negative bias at the control grid cuts off the flow 
of electron current through the pentode until the noise pulse has 
ended. Thus, no audio occurring during the period of a noise pulse 
will be reproduced at all. Fig. 3-4B shows the resultant output 
waveform of V2 when the noise pulses of Fig. 3-4A occur. Here, it 
can be observed that there is no output at all during the noise 
pulse. 

Since the noise pulse is of such short duration, lasting only for 
a portion of an audio cycle, or at the most, for just a few audio 
cycles, the absence is normally not noticeable to the listener. 
When the noise pulse passes, the electron pool which accumulated 

64 



on the right-hand plate of C2 will very quickly discharge to 
ground through LI, and the diode will again be able to conduct 
electrons on the succeeding positive half-cycles of audio voltage 
should another noise pulse occur. 

The statement appears frequently in the literature on noise- 
limiting and noise-cancellation circuits that the noise-limiting 
action punches a hole in the signal. This refers to the fact which 
is portrayed graphically in Fig. 3-4B, namely, that no audio output 
signal is delivered while a strong noise pulse is being received. 

LI and C2 act in much the same manner as a long time-constant 
RC filter. Noise pulses will occur at high frequencies, well above 
the audio range, so that the electrons in storage on C2 cannot dis- 
charge downward through LI after one noise cycle before the next 
such cycle occurs. 

SERIES-DIODE NOISE LIMITER 

The circuit shown in Figs. 3-5 and 3-6 is another popular noise- 
cancellation, or noise-limiting circuit. It derives its name from the 
fact that noise-limiting diode VI is in series with the audio signal 
path. During normal operation, the diode conducts continuously, 
and an output signal is developed across R5 for coupling to the 
pentode amplifier stage. During abnormal or noise-limiting opera- 
tion, the diode does not conduct so that no audio signal can be 
developed across R5 for the duration of the noise pulse. 

Identification of Components 

As far as possible, the components in Figs. 3-5, 3-6, and 3-7 have 
been labeled to coincide with their counterparts in the shunt noise 
limiter of Fig. 3-1, 3-2, and 3-3. The various components with their 
functional titles are as follows: 

Rl — Variable resistor (volume control). 

R2 — Voltage-dividing resistor for setting the noise level. 

R3 — Cathode biasing resistor. 

R4 — Noise-pulse filtering resistor. 

R5 — Diode load or output resistor. 

R6 — Grid-driving and grid-return resistor. 

CI — Audio storage capacitor. 

C2 — Blocking and coupling capacitor. 

C3 — Cathode bypass capacitor. 

C4 — Coupling and blocking capacitor. 

VI — Noise-limiting diode. 

V2 — Pentode audio-amplifier tube. 

V3 — Diode detector or demodulator. 

65 



PLATE 

CURRENT 

(MAXIMUM) 




BIASING 
CURRENT 



Fig. 3-2. Operation of the shunt-diode noise limiter — positive half -cycle. 

Identification of Currents 

The following electron currents will flow in this circuit during 
normal operation (meaning during conditions when no noise pulse 
is present) : 

1. Diode-detector load current, which is also the input audio 
signal (solid blue) . 

2. Voltage-divider current (solid green). 

3. Noise-pulse current (dotted blue). 

4. Diode current (dotted green) . 

5. Pentode grid-driving current (also in dotted green). 

6. Pentode-plate current (solid red). 

7. Cathode-filter current (also in solid red). 

During abnormal operation, meaning when an unwanted noise 
pulse is being received, the last four of the currents listed — the 
diode current, the grid-driving current, the pentode-plate current, 
and the cathode-filter current — do not flow. Since the plate cur- 
rent is invariably being used to deliver an audio signal to the next 
amplifier stage, or to some output device such as a speaker or 



66 



EXCESSIVE 

NOISE DIODE 

SIGNAL PLATE 

CURRENT 



NO 
PLATE 

CURRENT 




tv DIODE 
^BIASING 
CURRENT 



Fig. 3-3. Operation of the shunt-diode noise ii miter — excessive 
noise pulse being received. 

headphones, the cutting off of plate current during a noise pulse 
effectively cancels out any other adverse effects of that noise pulse. 

Details of Operation 

Fig. 3-5 depicts the current actions which occur during a nega- 
tive half -cycle of audio operation. The input current (shown 
in solid blue) is flowing at an audio frequency and represents the 
output of the V3 diode-detector circuit. This current is being 
driven downward through Rl, causing a flow into the left plate of 
C2, and downward through R4. As a result of this downward 
movement of electrons, the voltages at the tops of Rl and R4 will 
be negative. 

The variable tap on Rl enables any desired portion of this nega- 
tive voltage to be coupled to the diode noise limiter. This variable 
feature regulates the amount of electron current which is being 
driven into C2 and downward through R4; therefore it regulates 
the amount of negative voltage developed at the top of R4 during 
this negative half-cycle. 



67 



AUDIO OR 

MODULATION 

ENVELOPE 



NOISE PULSE— RELATIVELY FEW 
RF CYCLES OF EXCESSIVE 
AMPLITUDE 




RF waveform. 



NORMAL AUDIO SIGNAL 
AFTER DEMODULATION 



AUDIO SIGNAL OUTPUT FROM 

AMPLIFIER DROPS TO ZERO OR 

LOW VALUE DURING NOISE PULSE 




(B) Audio waveform after noise limiting. 
Fig, 3-4. Effect of noise pulses on the RF and audio waveform. 

VI is biased by the voltage existing at voltage divider R2 so that 
it conducts continuously during normal operation. This is accom- 
plished when the voltage divider current through R2 (shown in 
solid green) flows continuously in the counterclockwise direction, 
being drawn upward from ground and through R2 to the positive 
terminal of the power supply. The diode plate is connected to this 
positive terminal, whereas its cathode is connected to some point 
of lower positive voltage on R2. Because the plate is more positive 
than the cathode, the diode current shown in dotted green will 
also tend to flow continuously. Its complete path begins at the 
ground connection at the right-hand end of R2. It then flows 
through part of R2, upward through R5, through the diode, down- 



68 



ward through R4, into the positive terminal of the power supply, 
and through it to ground. 

Because of this diode current flow, the voltage at the plate of 
the diode will be less than the power supply voltage by the amount 
of current "dropped" or developed across R4 by this same current. 
Also, during a negative half-cycle of audio, such as is shown in 
Fig. 3-5, a negative voltage is developed across R4 by the input 
current (shown in solid blue) . The amount of this negative voltage 
must be subtracted from the positive voltage which exists at the 
plate because of the biasing actions just described. Therefore, dur- 
ing a negative half-cycle of audio, the positive voltage at the diode 
plate will be reduced. This will cause a reduction in the amount 
of diode current (shown in dotted green); this reduction causes 
a smaller voltage drop to exist across the diode cathode load R5. 

The voltage at the top of R5 and the cathode of the diode is posi- 
tive in polarity at all times. During noise-pulse reception (which 
is described later) when no diode current flows, the lowest positive 
voltage which the cathode can attain is reached; this will be the 
same amount of positive voltage as that which exists at the point 
on voltage divider R2 where the variable tap is placed. When a 
small amount of diode current flows during the negative half- 
cycles, the voltage at the top of R5 will be only slightly more posi- 
tive than this value. When a large amount of plate current flows 
during positive half -cycles, the voltage at the top of R5 will be con- 
siderably more positive than this value. 

Thus, it can be seen that the voltage at the diode cathode and the 
top of R5 fluctuates between two values of positive voltage, in 
accordance with the flow of audio-driving current up and down 
through R4. On negative half-cycles, such as are shown in Fig. 
3-5, the diode cathode voltage has its least positive value. On these 
half-cycles, electron current will flow into the left-hand plate of 
C4. This action will drive an equal number of electrons downward 
through grid-driving resistor R6. This is the grid-driving current 
for V2 (shown in dotted green). The downward flow of current 
through R6 places a negative voltage on the grid of V2, thereby 
reducing the plate current through this tube to a minimum value. 

During a positive half-cycle of audio, such as that shown in 
Fig. 3-6, the detector current flowing through V3 and Rl cannot 
reverse its direction, since a diode is a unidirectional device. How- 
ever, this current is reduced to a low value on positive half-cycles, 
causing a reversal of current direction in capacitors CI and C2 
and in resistor R4. During the negative half-cycles of Fig. 3-5, the 
upper plate of CI "fills up" with accumulated electrons. During 
the positive half -cycles of Fig. 3-6, this reservoir becomes 
depleted. 

69 



® 



It 




AMPLIFIER 

PLATE 
CURRENT 



_L 




)IODE ■ 
-BIASING 
^'CURRENT 



tm 



Fig. 3-5. Operation of the series-diode noise limiter — negative half-cycle. 



W 



ci)=S: 



INPUT 

AUDIO 

SIGNAL 

CURRENT 



NORMAL 

DIODE 
CURRENT 



u 




GRID 
DRIVING 
CURRENT 






i@ 



* 3 0WER ^» C 

I SUPPLY JT 



AMPLIFIER 

PLATE 
CURRENT 

JL. 



I 



DIODE 
BIASING 
CURRENT 



Fig. 3-6. Operation of the series-diode noise limiter — positive half-cycle. 
70 



NO DIODE 
CURRENT FLOW 
(TUBE CUTOFF] 



U£*2 1 




C4 CHARGING 
ACTION WHICH 
MAKES DIODE 
CATHODE MORE 
NEGATIVE WHEN 
TUBE IS CUTOFF 



DIODE 4r 
BIASING " 
"*? CURRENT 




Fig. 3-7. Operation of the series-diode noise limiter — strong 
noise pulse being received. 

The left-hand plate of C2 acts in a similar manner, accumulating 
electrons during negative half -cycles and discharging them back to 
Rl during positive half-cycles. In consonance with the charge 
and discharge action of C2, the input current (shown in solid blue) 
flows downward through R4 during negative half-cycles, and up- 
ward through R4 during the positive half -cycles. Thus, during the 
positive half-cycles, the voltage developed across R4 by this up- 
ward current flow will be positive at the top of R4, thus counter- 
acting, to some extent, the negative component of the voltage de- 
veloped across this same resistor by the continuous downward 
flow of current coming through diode VI. Therefore, the diode 
plate is made more positive, increasing the diode current flow. At 
these times (the positive half-cycles) the cathode of VI will reach 
its highest positive voltage; and on these half -cycles, electrons 
will be drawn out of the left-hand plate of C4 to supply the in- 
creased demand for electrons flowing into the diode. This action 
draws an equal number of electrons upward through resistor 
R6, and their upward flow (shown in dotted green in Fig. 3-6) 
makes the grid of pentode V2 positive, thereby causing the 
maximum amount of plate current (shown in solid red) to flow 
through V2. 



71 



Operation During Noise-Pulse Reception 

Fig, 3-4A indicates a typical noise pulse and the manner in 
which it will distort a normal carrier wave. Resistor R2 acts 
as a noise-level control in this circuit and will normally be set to a 
position so that the diode will conduct electrons during the entire 
range of any audio cycle which might be received. However, when 
a noise pulse is received, it is desirable that the diode not conduct. 
Fig. 3-7 shows the additional noise current that flows during re- 
ceipt of a noise pulse; from this illustration we can see how the 
noise current cuts off the diode and further results in cutting off 
audio-amplifier tube V2. 

The noise current (shown in dotted blue in Fig. 3-7) is flowing 
downward through R4. Since the noise signal is by nature much 
stronger in amplitude than the normal audio signal, this noise 
current develops a much stronger component of negative voltage 
at the top of R4. When this negative voltage is large enough, it will 
exceed the positive voltage which is simultaneously developed 
between the diode plate and cathode by the voltage-divider cur- 
rent flowing from right to left through R2. When this happens, the 
diode stops conducting. 

When the diode stops conducting, the cathode voltage drops to 
its lowest positive value; this voltage drop is "coupled" across C4 
to grid-driving resistor R6. Translated into terms of current flow, 
we find electrons flowing onto the left-hand plate of C4, driving 
other electrons downward through R6, and making the voltage at 
the top of R6 negative enough to cut off the electron flow through 
the pentode amplifier entirely. This portion of the action is identi- 
cal (except in degree) to that which is depicted in Fig. 3-5 for C4 
and R6, when a negative half-cycle of audio occurs. 

DUAL-DIODE NOISE LIMITER 

Figs. 3-8 and 3-9 show two separate moments in the operation 
of a dual-diode noise limiter. Fig. 3-10 shows a typical audio 
waveform, as distorted by a strong noise pulse. This particular 
circuit may be used in the audio section of a receiver to protect 
against a strong negative or positive noise pulse. 

Identification of Components 

The circuit of Figs. 3-8 and 3-9 is composed of the following 
components: 

Rl — Grid-driving resistor. 
R2 — Plate-load resistor. 

72 



R3 — Grid-driving resistor. 

R4 — Cathode-biasing resistor. 

CI — Coupling and blocking capacitor, 

C2 — Cathode-filter capacitor. 

VI — Triode audio amplifier. 

V2 — Positive limiter diode. 

V3 — Negative limiter diode. 

V4 — Triode audio amplifier. 

Ml — Bias battery (or other voltage source). 

M2 — Bias battery (or other voltage source). 

Identification of Currents 

The following currents are at work in this circuit during normal 
(no noise pulses) operation: 

1. Two grid driving currents (solid blue). 

2. Two triode plate currents (solid red). 

3. Cathode filter for V4 (dotted red). 

During abnormal operation when a noise pulse is present, one 
of the following additional currents will flow. 

4. Positive limiting diode current through V2 (dotted green) . 

5. Negative limiting diode current through V3 (solid green). 

(Both of these currents will flow if the noise pulse has both posi- 
tive and negative components, but the currents cannot flow simul- 
taneously — they must flow in sequence, or consecutively.) 

Details of Operation 

With the two diodes and bias batteries removed from the circuit, 
it becomes a conventional RC coupled audio amplifier and, in 
normal operation, acts like one. In Fig. 3-8, if you disregard the 
current which flows through V2 (dotted green) , you see what ap- 
pears to be a positive half -cycle in the operation of VI. (The 
term positive as applied to a half -cycle of operation is an arbitrary 
one and can be taken to refer either to the instantaneous grid 
voltage or the instantaneous plate voltage. It refers to the plate 
voltage of VI in this example.) 

The grid voltage is negative during this half-cycle, as shown by 
the downward movement of grid-driving electrons through Rl 
(solid blue). The negative grid voltage reduces the flow of plate 
current through VI and the downward flow of plate current 
through load resistor R2. Consequently, the voltage at the plate 
of VI must become more positive. This rise in plate voltage draws 
an electron current upward through grid resistor R3 so that the 

73 



NEGATIVE 
HALF- CYCLE 
OF GRID- 
DRIVING 
CURRENT 



PLATE 

VOLTAGE 

RISES 

2 



GRI D VOLTAG E RISES ™„ 

AND DIODE CONDUCTS ^?Jl^-r 

IF GRID VOLTAGE 
BECOMES TOO POSITIVE 




Fig. 3-8. Operation of the dual-diode noise limiter — positive 
noise pulse being received. 



top of the resistor becomes positive in voltage, (The grid-driving 
current in R3 has also been shown in solid blue.) 

In Fig. 3-10, that portion of the audio waveform which is above 
the center line has been arbitrarily labeled as "positive." That 
portion of the waveform which appears sinusoidal in nature is 
considered to be within the normal operating limits of the circuit. 
The bias battery, or voltage source shown below the cathode of 
V2 must be chosen so that it is equal to or greater than the voltage 
represented by the normal operating limit voltage. When this is 



POSITIVE 
HALF- CYCLE 
OF GRID- 
DRIVING 
CURRENT 




PLATE 
VOLTAGE 
FALLS 

-A 



GRID VOLTAGE BECOMES GRID 

NEGATIVE AND DIODE V3 DRIVING 

CONDUCTS IF GRID0FV4 CURRENT 
BECOMES TOO NEGATIVE 




Fig. 3-9. Operation of the dual -diode noise limiter — negative 
noise pulse being received. 



74 



done, under normal conditions, V2 cannot conduct because the 
diode plate voltage can never become more positive than the 
cathode voltage. 

When a strong noise pulse having a positive polarity is received, 
normal conditions are exceeded. The most positive voltage that the 
plate of VI can attain is the value of plate-supply voltage provided 
by the power supply. It will reach this value only if and when the 
control grid of VI is made negative enough to cut off the flow of 
plate current entirely. A strong noise pulse would be the most 
likely cause of VI cutting off. 

Before the plate of VI (and the grid of V2) can become this 
positive, the plate of V2 will become more positive than its cath- 
ode, with the result that diode current will flow. This current 
(shown in dotted green) flows along the path shown from cathode 
to plate within V2 and downward through R3 to ground. Since it is 
flowing through R3 in a direction opposite to the flow of the grid- 
driving current, it partially neutralizes the high positive voltage 
which would otherwise be developed across R3. 



POSITIVE NOISE PULSE 



POSTIVE HALF 
CYCLE OF NORMAL 
AUDIO 




NEGATIVE HALF 
CYCLE OF NORMAL 
AUDIO 



NEGATIVE NOISE PULSE 



Fig. 3-10. Amplitude relationships between normal 
audio cycles and unwanted noise pulses. 



75 



As the noise pulse increases in strength, it tends to draw more 
electron current upward through R3; this, in turn, makes the 
diode plate even more positive and thereby increases the amount 
of diode current. Thus, a larger amount of diode current is avail- 
able to flow downward through R3, tending to neutralize any 
increased positive voltage at the grid of V4 due to a stronger noise 
pulse. 

Fig. 3-9 shows what might be called a negative half-cycle in 
the operation of the circuit (the term "negative" being used to 
describe the direction in which the plate voltage of VI changes). 
The lowest point on the audio sine wave of Fig. 3-10 is well within 
the operating limits of the circuit; no current will flow through 
diode V3. The grid of VI is made positive during this half -cycle 
as a result of current being drawn upward through Rl. This re- 
leases a large pulsation of plate current into tube VI, and by virtue 
of their downward flow through load resistor R2, the plate voltage 
is lowered. This drop in voltage at the plate of VI drives electron 
current downward through grid resistor R3, making the grid of 
V4 negative. 

If a strong negative noise pulse is being received, the voltage at 
the grid of VI will be made excessively positive, probably causing 
"saturation" current to flow in VI. This would lower the plate 
voltage of VI to its lowest possible value, and thus, in turn, drive 
the grid of V4 so negative that it would probably be cut off. This 
is an undesirable condition, and one which exceeds the normal op- 
erating limits of the circuit. The bias voltage below the plate of 
diode V3 has a value which is so chosen as to be considerably less 
than this value of cutoff grid voltage. Consequently, before the 
grid voltage can reach such a negative voltage, the cathode of V3 
will become more negative than its plate, and V3 will conduct an 
overload, or limiting current (shown in green). It flows upward 
through R3 and downward from the cathode to the plate of diode 
V3. 

Since it flows upward through R3 while the grid-driving current 
is flowing downward through the same resistor, it partially offsets, 
or neutralizes, the large negative voltage which would otherwise 
exist at the grid as a result of the negative noise pulse. If the noise 
pulse becomes stronger, the diode conducts more electron current 
and thus tends to counteract the effects of the noise pulse flowing 
through R3. 

Since both the positive and negative noise pulses cannot be 
occurring simultaneously (or they would cancel each other out) 
the two noise limiting diodes cannot conduct at the same moment. 
A noise pulse may be either positive or negative at any given 
instant, but not both. 

76 



SQUELCH CIRCUIT 

In the reception of certain types of communications, it is 
necessary for someone to be listening to the receiver at all times, 
even when no signal is being received. This is done so that when 
a signal does come in, it will not be missed. A Federal airways 
ground station guarding several different channels or frequencies 
on several different receivers, all simultaneously, is a good 
example of this. There is a certain disagreeable background noise 
or hissing which comes from a receiver under conditions of no- 
signal reception. With two or more channels being guarded at the 
same time, this combination of background noises becomes most 
unpleasant, often leading even to inattentiveness on the operator's 
part and ultimate loss in communications. 

The squelch circuit is a simple combination of parts which is 
designed to eliminate this undesirable condition. The squelch 
circuit is a carrier-operated device, or switch, which turns the 
audio amplifier off when no carrier signal is being received and 
turns it on when a carrier signal is present. The switching action 
uses the negative voltage which is associated with the AVC action 
to accomplish this function. 

Two identical circuit diagrams have been selected to illustrate 
how this function has been accomplished. Fig. 3-11 shows a 
typical half-cycle of audio operation when a signal is being re- 
ceived. Fig. 3-12 depicts operating conditions when no signal is 
being received and the squelch circuit cuts off the amplifier tube. 

Identification of Components 

The squelch circuit and the audio amplifier contain the follow- 
ing individual components, with functions as indicated: 

Rl — Isolating resistor. 

R2 — Grid-return resistor for squelch tube, 
R3 — Variable resistor used as Squelch Control. 
R4 — Load resistor for VI; also serves as grid-driving and grid- 
return resistor for V2. 
R5 — Additional grid-return resistor. 
R6 — Voltage-divider resistor. 
R7 — Cathode-biasing resistor. 
R8 — Amplifier-load resistor. 
CI — AVC storage capacitor. 
C2 — Screen-filtering capacitor. 
C3 — Audio-input coupling capacitor. 
C4 — Cathode-bypass capacitor. 
C5 — Output-coupling capacitor. 

77 



INPUT 
AUDIO 

SIGNAL 
CURRENT 



NEGATIVE AVC 

VOLTAGE (BIASES 

SQUELCH TUBE 

TO CUTOFF} 




POWER 
SUPPLY 



CATHODE 
FILTERING 
CURRENT 



VOLTAGE DIVIDER 

AND BIASING 

CURRENT 

Fig. 3-11. Operation of the squelch circuit — negative half -cycle of 
audio being received. 

VI — Pentode used for squelching purposes. 
V2 — Triode audio amplifier. 

Identification of Currents 

During the normal operation, when a carrier signal is being re- 
ceived, the following electron currents will flow: 

1. Current discharge from AVC storage capacitor (solid green) . 

2. Input audio current (solid blue). 

3. Voltage-divider current (dotted green). 

4. Amplifier-plate current (solid red). 

5. Cathode-filter current (also in solid red). 

When no carrier signal is present, the voltage divider current 
(shown in dotted green) is the only one of the foregoing currents 
that will flow. However, plate and screen current (both shown 
in solid red) will then flow in squelch tube VI. 

Details of Operation 

When a carrier signal is being received in a typical receiver, it 
will normally be used, among other things, to generate an AVC 
voltage. This voltage is stored on a large- value AVC storage 
capacitor. CI in Figs. 3-11 and 3-12 provides this function. 



78 



NO AVC VOLTAGE 

ON CI BECAUSE 

NO SIGNAL IS 

BEING RECEIVED 





VOLTAGE DIVIDER 

AND BIASING 

CURRENT 

Fig. 3-12. Operation of the squelch circuit — no audio being received. 

The AVC voltage is a negative voltage, consisting of an accum- 
ulation of electrons in storage on one plate of the capacitor. When 
the received carrier signal increases in strength, additional elec- 
trons are placed in storage on the capacitor, thereby increasing 
the amount of the negative AVC voltage. When the received 
carrier signal is reduced in strength, some electrons are given up 
from storage, thereby reducing the amount of the negative AVC 
voltage. When no carrier signal is present, all electrons in storage 
on the AVC capacitor will discharge to ground through Rl and 
R2 (Fig. 3-12) . 

None of the circuitry necessary for the generation of the AVC 
voltage has been shown in Figs. 3-11 and 3-12, nor discussed in 
this chapter. Refer to Chapter 2 of this book or to an earlier text, 
Detector and Rectifier Circuit Actions, for a full discussion of this 
circuitry. 

The negative voltage stored on the upper plate of CI causes a 
continual discharge of electrons through Rl and R2, as shown in 
Fig. 3-11. This places the grid of squelch tube VI at a sufficiently 
negative voltage to cut off the normal flow of plate current through 
this tube. Thus, whenever a carrier signal is being received, the 
squelch tube does not conduct. Under these conditions V2 is biased 
by the voltage divider current (shown in dotted green) as well as 



79 



by its own cathode-to-plate current (shown in solid red). It is 
driven by the audio-input signal (shown in solid blue). 

The voltage-divider current flows continuously along the path 
shown, upward through R3 and downward through R6 and into 
the power supply. Thus, a fixed positive voltage is provided at the 
junction of these two resistors. It will be noted that both the cath- 
ode and the control grid of V2 are returned to this same point, 
therefore the positive voltage resulting from the flow of voltage 
divider current biases the cathode and grid of V2 equally. This 
intrinsic voltage could be almost any value up to full B+, de- 
pending on the choice of sizes for R3 and R6. 

When plate current flows through V2, it must follow the com- 
plete path indicated in solid red in Fig. 3-11. This current origi- 
nates at ground below R3, and flows upward through R3 before 
entering cathode resistor R7 and flowing upward through it to 
the cathode. It then flows through the tube from cathode to plate, 
downward through R8, and into the power supply, then through 
it to the common ground where it has ready return access to R3. 
In flowing upward through R7, an even more positive voltage is 
created at the cathode. Thus, the cathode is biased positively with 
respect to control grid, or stated differently the control grid is 
biased negatively with respect to the cathode — the normal operat- 
ing condition for an amplifier. 

The audio-input current (shown in solid blue) is coupled to the 
amplifier circuit via C3. Its complete path is, into the left-hand 
plate of C3, driving an equal number of electrons out of the right- 
hand plate and downward through R4 and R3 to ground. These 
two resistors, R4 and R3, consequently serve as grid-driving re- 
sistors for V2 in addition to their other functions as a plate-load 
resistor (R4) for VI and a biasing resistor (R3) for the screen 
grid of VI and the control grid and cathode of V2. 

During the negative half -cycle of operation depicted in Fig. 3-11, 
the control grid of V2 will be made more negative than usual, and 
the plate current flowing through the tube will be reduced to a 
low value. Also, at this time, the cathode filter current for V2 
will be flowing downward to filter capacitor C4. 

During the positive half-cycles of audio operation, the audio 
current and the cathode-filter current will reverse directions. 
That is, audio current will be drawn out of C3; this action will, in 
turn, draw an equal number of electrons upward through R3 and 
R4 and into the right-hand plate of C3. This action adds a 
component of positive voltage at the control grid of V2, or, stated 
differently, it reduces the amount of negative bias existing be- 
tween the grid and cathode of V2. Thus, more plate current will 
flow through the tube during the positive half cycles. 

80 



Operation When No Carrier Signal Is Present 

Fig. 3-12 depicts the two currents which flow in the squelch 
circuit when no carrier signal is being received. When there is no 
carrier signal, there is, of course, no audio signal to be demodu- 
lated, and no AVC voltage can be developed from it. 

When the negative AVC voltage normally stored on CI is re- 
moved, VI will begin to conduct. This plate current begins at the 
ground connection below the cathode, passes through the tube, 
through R4 and R6, through the power supply to ground. 

This downward flow of electron current through R4 makes the 
upper end of this resistor negative with respect to the lower end. 
Since the grid of V2 is connected to the upper end of R4, while 
the cathode of V2 is connected to the lower end of R4 t conditions 
are suitable for restricting or cutting off the flow of plate current 
through V2. Normally, the amount of plate current through VI 
will be of sufficient magnitude that it develops a large enough 
component of negative voltage across R4 to cut off V2. This, of 
course, is the basic purpose of a squelch circuit — to prevent the 
audio amplifier from conducting when no carrier is being received. 

The variable tap on squelch control R3 can be used to vary the 
amount of positive voltage on the screen grid of VI. The squelch 
control provides a simple means of selecting or rejecting signals 
of any desired strength or intensity. This feature is particularly 
attractive for adapting to variable atmospheric phenomena, as 
well as varying operating conditions or criteria. 

REVIEW QUESTIONS 

1. In the circuit of Fig. 3-1, what may be denned as reception of a 
circuit action determines the bias signal of normal strength, with 
yoltage of diode VI? no noise pulses present.) 

2. In this same circuit, what circuit 6. What is the fundamental differ- 
action causes VI to conduct? ence in the operation of the dual- 

3. When VI conducts, what effect diode limiter of this chapter as 
does it have on the voltage at the compared to the two single-diode 
control grid of amplifier tube V2 limiters previously discussed? 
and why? Does this difference make the 

4. Does this same noise-limiter cir- dual-diode circuit more or less 
cuit operate on negative noise versatile than the single-diode 
pulses, on positive noise pulses, circuits. 

or on both? "- What is the main reason for 

5. In the circuit of Fig. 3-5 is the usin ? * squelch circuit in a corn- 
diode normally "conducting" or munications receiver? 

"cut off"? How does this compare 8. Will squelch tube VI in Figs, 
with normal operation of the 3-11 and 3-12 be caused to con- 
shunt diode limiter previously duct or to be cut off by the exist- 
discussed? (Normal operation ence of an AVC or AGC voltage? 



81 



Chapter 4 

HALF-WAVE POWER SUPPLY 



The primary function of a rectifier power -supply circuit in a 
typical piece of electronic equipment, such as a radio, is to convert 
the alternating current (AC) which is supplied to homes to direct 
current (DC). This is a necessary function, because vacuum-tube 
circuits require the application of fairly high and stable voltages 
to the tube plates and screens. The purpose of this chapter is to 
clarify the difference between an alternating current, or voltage, 
and a direct current, or voltage, and then to show how this direct 
voltage is achieved with this circuit. 

HALF-WAVE RECTIFIER CIRCUIT 

Figs. 4-1 and 4-2 show the two half-cycles of operation in a half- 
wave rectifier circuit. While a transformer has been employed in 
this circuit, it is often omitted and the power line is connected 
(through the switch) directly to the rectifier tube plate. In either 
case, operation of the circuit is the same. 

Identification of Components 

The components of the power supply circuit shown in Figs. 4-1 
and 4-2 and their function are as follows: 

RIO — Filter resistor. 

C14 — Filter capacitor. 

C15 — Filter capacitor. 

T4 — Power transformer. 

V5 — Half -wave rectifier tube. 

82 



ON -OFF 
SWITCH 



INSTANTANEOUS 
TRANSFORMER 
POLARITIES ' 




60-CYCLE CURRENT 
FROM HOUSE SUPPLY 



-I 



i r\ A 



+ © 



F1L 
(W) (V2) (V3) 



60-CYCLE ^f 
FILTER CURRENTS 



TRANSFORMER 
SECONDARY 
CURRENT 
(FLOWS THROUGH 
ALL FILAMENTS TO 
HEAT THEM) 




COMMON 
GROUND 



Fig. 4-1. Operation of a half-wave power supply — positive half cycle. 

In addition, the filament circuits for all tubes in the radio (VI 
through V5) are included in Figs. 4-1 and 4-2. Each of these 
filaments heats the cathode of its respective tube so that the 
cathodes may emit electrons within the tubes. 

Identification of Currents 

There are four separate and distinct electron currents at work in 
the circuit of Figs. 4-1 and 4-2 during normal operation. These 
electron currents are: 

1. Transformer primary current (solid blue). 

2. Transformer secondary current (dotted blue). 

3. Rectifier plate current (solid red). 

4. 60-cycle filter currents (dotted red). 

Details of Operation 

The On-Off switch of the radio is shown in the upper left-hand 
corner of Figs. 4-1 and 4-2. When this switch is opened, or "Off," 
the radio is isolated from the house electrical supply, and no 
electron currents will flow in any of the circuit components. 
When the switch is closed, or "On," (Figs. 4-1 and 4-2) the house 
current (solid blue) will flow back and forth through the primary 
winding of T4. 



83 



A half-cycle when the house current is flowing upward through 
the primary winding of T4 is depicted in Fig. 4-1. The instan- 
taneous voltage polarities across the primary winding are as 
shown, namely, the upper end of the winding is negative and the 
lower positive. These are the polarities of the applied voltage, 
meaning that these voltages are applied from the house supply 
and, in turn, cause the primary current to flow. 

The secondary current is shown as flowing downward through 
the secondary winding. Associated with this current are the 
instantaneous polarities of the induced voltage, which are positive 
on the upper end and negative at the lower end of the secondary. 
This current also flows through the filaments of the five vacuum 
tubes, which are connected in series. During the half-cycle repre- 
sented by Fig. 4-1, this filament current flows to the left through 
all of the filaments. Each of the filaments is heated by this proc- 
ess very much as the coil of an electric toaster is heated. 

During the half-cycle represented by Fig. 4-2, the direction of 
flow of both the primary and secondary currents is changed. The 
primary current flows downward, and the secondary current 
flows upward and to the right through the tube filaments. The 
voltage polarities across the primary and secondary windings are 
also reversed during this half-cycle. Since we have assumed the 
supply current to be conventional 60-cycle, the positive and nega- 
tive half -cycles shown in these two illustrations are repeated 60 
times each second. 

Diode Tube Operation 

Referring again to Fig. 4-1, the diode plate current (solid red) 
is shown flowing from the cathode to the plate of V5. The complete 
path of this current is from the filter system made up of resistor 
RIO and capacitors C14 and C15, through V5, and through the 
secondary winding of the T4 to ground. 

In order for a vacuum tube, such as V5, to allow electron cur- 
rent to flow across the open space within the tube, two essential 
conditions must be met: 

1. The cathode must be heated by some external means so that 
it will emit electrons into the tube. This heating is accom- 
plished by placing the cathode very close to the filament, 
which is heated by the flow of transformer secondary current 
described previously. The emission process is analogous to 
the action that occurs at the surface of boiling water, when 
very small droplets of water appear to jump free from the 
surface. 

2. The plate of the tube must be at a more positive potential 

84 



than the cathode in order to attract the negative electrons 
across the tube. Since every electron possesses one unit of 
negative charge, it will be repelled by any negative voltage 
and attracted by any positive voltage. 

In V5 the first of these conditions is fulfilled whenever the 
On-Off switch is closed. The second condition is fulfilled only 
during the positive half-cycles depicted in Fig. 4-1. Thus, diode 
current (solid red) flows only during the positive half-cycles. 

It is important to understand the complete diode-current path. 
All of the plate and screen-grid currents from the four other tubes 
eventually join together at the junction of RIO and C15 to make 
up the diode current. This current flows through diode V5 on 
positive half -cycles only; therefore, it can be described as a pulsat- 
ing DC. The purpose of the filter system composed of RIO, C14, 
and C15, is to smooth out this pulsating flow of electron current so 
that the currents flowing from the other tube plates and screen 
grids will flow smoothly up to the filter circuit. 

The Filter Circuit 

RIO and C15 form a simple RC filter circuit, which operates to 
create a high positive voltage on the upper plate of C15. This high 
positive voltage is obtained as a result of the departure of many 
electrons which are drawn into V5. Electrons leave the upper plate 
of C15 in pulsations (whenever the plate of V5 is more positive 
than the cathode), creating an electron deficiency on the upper 
plate of C15. It is this electron deficiency, which is the same thing 
as a positive voltage, that attracts electrons from the plates and 
screen grids of other tubes. 

As electrons flow in from the other tubes, they would tend to 
equalize or neutralize the positive voltage on C15. However, other 
electrons are being drawn into V5 as fast as they are arriving 
from the other tubes, with the result that a positive voltage re- 
mains on the upper plate of C15 as long as the radio is turned on. 
Thus, we have a situation where electrons continuously flow onto 
the upper plate of C15, and intermittently flow from this same 
capacitor and into the diode tube V5. 

The amounts of electrons involved in these two current patterns 
will eventually stabilize and be equal to each other, resulting in a 
voltage on the upper plate of C15 which is almost as positive as 
the peak voltage on the upper terminal of the transformer sec- 
ondary winding. 

It is the positive voltage at the upper end of the transformer 
secondary winding that attracts electrons across V5 during the 
positive half -cycles. Likewise, it is the resulting positive voltage 



on the upper plate of C15 that draws electrons from the plates and 
screen grids of the other tubes. The positive voltage on the upper 
plate of C15 is a DC voltage, as a result of the filter action which 
occurs between RIO and C15. These two components form what is 
known as a long time-constant circuit. This is a mathematical 
term, but its significance can be explained and understood with 
the use of some simple arithmetic. When any resistor and capacitor 
are connected together, they have a time constant, which is de- 
termined by the product of the values of the two components. This 
relationship is expressed by the formula: 

T = Rx C 

where, 

T is the time constant of the combination in seconds, 
R is the resistor value in ohms, 
C is the capacitor values in farads. 

The time constant of any circuit is considered to be long when it 
is at least several times longer than the time or period of one cycle 
of the current which is passing through the combination. Assum- 
ing the value of RIO is 680 ohms and CIS is 80 mfd, the time con- 
stant of the two components is: 

T = Rx C 
= 680 x 80 x 10" 6 
= 54,400 x 10-° 
= .0544 second 
= approximately l/18th of a second. 

Since the frequency of the current this filter is trying to handle 
is 60 cps, the time or period of one of these cycles is one sixtieth 
of a second. Thus, the time constant of the combination of R10 and 
C15 is more than three times as long as one of these periods. 

The positive voltage on the upper plate of C15 can be likened 
to a pool of positive ions. As more negative electrons are drawn 
away from this pool and flow through V5, the number of positive 
ions on C15 will increase by the same amount. Also, as electrons 
flow into this pool of ions from the plates and screen grids of the 
other tubes. (Fig. 4-2) , the number of positive ions will be reduced 
accordingly. 

Positive ions in concentration represent a positive voltage the 
amount of which is directly proportional to the number of ions 
present. The voltage at the cathode of V5 and at the upper plate 
of C 15 in a typical radio will be +125 volts. This voltage does not 
change significantly from half -cycle to half -cycle, because the 
quantity of electrons which leave the capacitor on the positive 
half -cycles is an insignificant percentage of the quantity of positive 

86 



INSTANTANEOUS 
TRANSFORMER 
POLARITIES £ 




60-CYCLE CURRENT J^ 

FROM HOUSE SUPPLY 

COMMON 
GROUND 



LAMP 

CURRENT 



5> y\. 

-= 60-CYCLE -^ 
FILTER CURRENTS 

.® @ (^ @ 

* * a • • * A*** A»»*A*** # 



TRANSFORMER 
SECONDARY 
CURRENT 

(FLOWS THROUGH 
ALL FILAMENTS TO 
HEAT THEM) 



50C5 I2BE6 „,,„ 

I2BA6 I2AV6 



COMMON 
GROUND 



Fig. 4-2. Operation of a half-wave power supply — negative half cycle. 

ions already stored there, and the electrons which do leave are 
replenished during the negative half -cycles. 

Ripple Voltage 

All power-supply filter circuits exhibit a ripple voltage. The 
ripple voltage is the minute fluctuations in output voltage that 
exist as a result of electrons being drawn away from the voltage 
pool on C15 on the positive half-cycles. In a very long time- 
constant circuit, this ripple voltage will be an exceedingly small 
fraction of a volt. In a less sophisticated system, such as the 
typical radio, it will be much larger. However, if the ripple 
factor becomes too large, the operation of a radio will be ad- 
versely affected, and you will hear a 60-cycle hum along with 
the regular program. 

A filter capacitor, such as C15, operates like a mechanical shock 
absorber. On each positive half-cycle, such as is depicted in Fig. 
4-1, electrons are drawn away from the upper plate of C15 and 
through V5; an equal number will also be drawn upward from 
ground onto the lower plate of the capacitor. Likewise, on the 
negative half -cycles, such as are depicted in Fig. 4-2, V5 is not 
conducting, but the positive voltage on the upper plate of C15 will 
continue to draw electrons onto it from the plates and screen 
grids of all other tubes; this action will drive an equal number of 
filter current electrons downward from the lower plate of C15 to 



87 



ground. This filter current has been shown in dotted red in Figs. 
4-1 and 4-2. 

Other Tube Currents 

The electron currents which flow in and out of C14 are regu- 
lated and controlled by the same events which drive the currents 
in and out of C15, but to a lesser degree. All of the tube currents 
(plate and screen grid currents) from VI, V2, and V3, plus the 
screen-grid current from V4, must flow through RIO on their way 
to the power supply. (The plate of V4 is connected to the junction 
of C15 and RIO.) These currents flow toward V5 because the 
diode plate is made positive on the positive half-cycles drawing 
electron current from the upper plates of both C15 and C14. Fig. 
4-1 depicts these positive half-cycles, and shows electron current 
being drawn out of the upper plate of capacitor C14, and to the 
left through RIO to the cathode of the power supply diode V5. 

Fig. 4-2 depicts the negative half -cycle operation, when the 
diode plate is negative, and no electron current crosses V5. During 
these half -cycles, the positive voltage on the upper plates of C15 
and C14 will continue to draw electron current from the plates 
and screen grid currents of the other four tubes. During this half- 
cycle, these electron currents will flow onto the upper plates of 
these two capacitors, and will drive filter currents from the lower 
plates to ground. These flow directions have been indicated by 
arrows in Fig. 4-2. 



REVIEW QUESTIONS 



1. What are the two conditions 
which must be met in order for a 
diode to conduct? 

2. Describe the electron action (or 
actions) which keeps the upper 
plate of capacitor CIS in Fig. 4-1 
at a positive voltage. 

3. Inasmuch as electrons are drawn 
away from the upper plate of CI 5 
during a small portion of each 
cycle, why does the voltage on 



C15 not change appreciably dur- 
ing each cycle? 

In Figs. 4-1 and 4-2 electron cur- 
rent shown in solid red is flowing 
from right to left through resis- 
tor RIO. What is this current 
usually called and where does it 
come from? 

In the 35 W4 tube of this chapter, 
does the filament heating current 
shown in solid blue cross the di- 
ode tube from cathode to plate? 



88 



Chapter 5 

REGENERATION 



Regenerative circuits were widely used in the early days of 
radio. For example, a regenerative detector was often employed to 
detect, or demodulate, very weak radio-frequency signals be- 
cause of its ability to amplify the signal as it was demodulated. 
Such signals might have been continuous wave (CW) signals used 
for the transmission of code, or amplitude-modulated (AM) 
signals carrying voice or entertainment-type information. 

REGENERATIVE DETECTOR 

A typical regenerative detector circuit is given in Figs. 5-1, 5-2, 
and 5-3. Gains in signal strength of 10,000 or 12,000 are common 
with this type of circuits. In modern entertainment-type equip- 
ment, the regenerative detector is seldom encountered; however, 
it is often employed in small communications-type receivers. 

The development of new RF amplification techniques has led to 
the abandonment of the regenerative detector in entertainment- 
type equipment. For example, an RF amplifier circuit with a volt- 
age gain, or amplification, of 30 times is not uncommon. Three 
such stages in series, or cascade, would have an over-all gain of 
27,000. While such a circuit would require three tubes instead of 
one and appear to invite more complexity and cost, it eliminates 
the inherent disadvantages of the regenerative detector, namely, 
the need for critical adjustment. A slight misadjustment in a 
"regen" detector can cause it to go into self-sustained oscillation, 
even in the absence of a radio-frequency signal. 



NORMAL PLATE CURRENT 
WHEN NO SIGNAL IS 
PRESENT (PURE DC) 




VVVV I 

• — 

▼1 



® 



VOLTAGE-DIVIDER 
CURRENT 



Fig. 5-1. Operation of the regenerative detector — no received signal. 

PLATE CURRENT 



®RF CHOKE 
^S CURRENT 



HEADPHONE 
CURRENT 




FEEDBACK 

CURRENT GR| D-I A NK 
CURRENT 



GRID-DRIVING CURRENT 
(MAKES GRID LESS 
NEGATIVE) 



' WW i 

. . f **»• • • • • • I 

/ :i 

• "ST 




® 



VOLTAGE- DIVIDER 
CURRENT 



Fig. 5-2. Operation of the regenerative detector — positive 
half -cycle of signal. 



90 



PLATE CURRENT 
(PULSATING DC) 

L 



HEADPHONE 
CURRENT 




FEEDBACK 

CURRENT s *!£l™|. K 
CURRENT 



GRID-DRIVING CURRENT 
(MAKES GRID MORE 
NEGATIVE 



VOLTAGE- DIVIDER 
CURRENT 

Fig. 5-3. Operation of the regenerative detector — negative 
half-cycle of signal. 



Identification of Components 

The components which make up the regenerative detector 
circuit are shown in Fig. 5-1, 5-2, and 5-3, and their functions are 
as follows: 

Rl — Grid-leak biasing resistor. 

R2 — Variable resistor used as voltage divider. 

CI — Tuned-tank capacitor. 

C2 — Coupling and biasing capacitor. 

C3 — RF filter capacitor. 

LI — Primary winding of RF transformer. 

L2 — Secondary winding of RF transformer. 

L3 — Tickler or regeneration coil. 

L4 — Radio-frequency choke coil. 

L5 — Primary winding of AF transformer. 

L6 — Secondary winding of AF transformer. 

VI — Detector-amplifier tube. 

Ml — Power supply. 

M2 — Headphones. 



91 



Identification of Currents 

In the absence of an RF signal, only two significant electron 
currents will flow in this circuit. These currents are: 

1. Plate current through VI (solid red) . 

2. Voltage-divider current through R2 (dotted red). 

When an RF signal is being received, seven additional currents 
flow in the circuit. These currents are as follows: 

3. RF input or signal current (dotted blue). 

4. RF tank current (solid blue). 

5. RF grid-driving current (also in solid blue). 

6. Grid-leakage current (dotted green). 

7. RF filter current through C3 (also in solid blue). 

8. Audio current through headphones (solid green). 

9. Feedback, or regenerative, current (also in dotted red). 

Details of Operation 

In Fig. 5-1, when no RF signal is being received, only two elec- 
tron currents — the voltage-divider and the tube current— are 
flowing. The voltage divider is placed across the power supply to 
provide the operator a means of controlling the amount of regen- 
eration and, consequently, the amount of amplification available 
from the tube. The voltage-divider current (shown in dotted red) 
flows continuously from ground, from left to right through R2 
and enters the positive terminal of the power supply, returning 
through the power supply to ground. Because of this continuous 
current movement through R2, a progressively higher positive 
voltage exists at each successive point from left to right. Thus, the 
position of the movable arm on R2 determines the amount of posi- 
tive voltage applied to the plate of VI; this directly affects the 
amount of tube current that will flow. 

The tube current (shown in solid red) consists of electrons 
which are drawn out of ground below the cathode. The heated 
cathode causes them to be emitted into the tube where they are 
attracted across the tube by the positive voltage on the plate. From 
the plate, they flow successively through L3, L4, L5, and R2. 
From R2 they are drawn into the positive terminal of the power 
supply, through which they must be delivered back to ground in 
order to have ready return access to the cathode of the tube. 

Once stable conditions exist and when no signal is being re- 
ceived, this plate current is a pure DC. Consequently, no feedback 
or regeneration can exist between L3 and L2, or L5 and L6. 

92 



In Fig. 5-2 the additional currents which come into existence 
when a signal is being received from some transmitting station are 
shown. LI may be considered as connected directly to an antenna; 
therefore the signal current induced in the antenna flows directly 
up and down through LI. This current is shown in dotted blue; 
in Fig. 5-2, it is flowing upward through the coil. The continual up 
and down flow of this current through LI induces a companion 
current to flow down and up through secondary winding L2, 
This current is shown in solid blue; in Fig. 5-2 it is flowing 
downward. This action delivers electrons to the lower plate of CI, 
making it negative, and withdraws electrons from the upper plate 
of CI, making it positive. Whenever the voltage on the upper 
plate of CI is positive, it draws electrons toward it from any 
external circuit to which it may be connected. In this circuit it 
draws electrons upward through grid resistor Rl and onto the 
left-hand plate of C2. This current (shown in solid blue) becomes 
the electron current which drives the grid; in Fig. 5-2 it drives the 
grid to a positive peak of voltage. 

(A simple rationale for correctly relating current directions to 
the resulting voltage polarities exists. Since electrons are them- 
selves negative in nature, they will always flow away from a more 
negative area and toward a less negative area. Thus, an upward 
flow of these grid-driving electrons through Rl tells us that the 
voltage at the upper end of the resistor is more positive than that 
at the lower end.) 

When the grid of VI is made positive by this flow of grid-driving 
current, two important things happen within the tube: first, the 
amount of plate current flowing through the tube is increased; 
and second, grid leakage electrons flow out of the tube at the con- 
trol grid. 

The additional surge of plate current must flow upward through 
L3 on its way to the power supply. As this plate current increases, 
it induces a separate current in L2, to which coil L3 is inductively 
coupled. This new induced current, which is shown in dotted red, 
is the feedback current which provides the basic regenerative ac- 
tion which gives the circuit its name. Since it is flowing down- 
ward simultaneously with the downward flow of the tank current, 
the feedback current reinforces the tank current. 

Fig. 5-3 shows the current conditions a half-cycle later. It is 
called a negative half-cycle because the grid-driving current is 
flowing downward through Rl, making the voltage at the top of 
this resistor negative. This grid-driving current is itself being 
driven by the negative voltage on the upper plate of tank capacitor 
CI. This negative voltage results from the fact that the tank- 
current electron flow has reversed during this half-cycle, and 

93 



electrons are flowing upward through L2 to the upper plate of CI, 
making it negative. 

The negative voltage at the control grid during this half-cycle 
restricts or reduces the plate current flowing through the tube. 
A decrease in the amount of current flowing upward through L3 
can induce a decrease in the current flowing downward through 
L2, or it can induce an increase in a current flowing upward 
through L2. For convenience, the latter case has been depicted in 
Fig. 5-3. Since the feedback and tank currents are both flowing 
upward simultaneously, one reinforces the other, and regeneration 
of the received signal occurs on both half -cycles. 

Because of regeneration, an extremely weak input signal flow- 
ing through LI may be amplified many thousands of times. With 
the appropriate adjustment of R2 (which controls the amount of 
plate current through the tube), the circuit can be operated just 
below the point of oscillation. A slight increase in the coupling 
between L2 and L3, or in the plate voltage applied to the tube, 
would cause the circuit to go into self -sustained oscillations, even 
in the absence of an input signal. This adjustment is one of the 
major disadvantages of the regenerative detector. 

On the positive half -cycles depicted by Fig. 5-2, grid-leakage 
current flows. The complete path of this current (shown in dotted 
green) begins as usual at the ground connection below the cath- 
ode of the tube. Whenever a control grid has positive voltage on it, 
it will attract some of the electrons from the electron stream pass- 
ing through the tube. Once these electrons strike the control-grid 
wires, they cannot be re-emitted into the tube; therefore, they 
must exit via the control grid. Eventually, the grid-leakage elec- 
trons will flow downward through grid resistor Rl and back to 
ground. If Rl is made large enough (as it usually is in practical 
circuit design), the electrons cannot flow immediately to ground 
but will first accumulate on the right-hand plate of C2, thereby 
building up a negative voltage, known as a grid-leak bias voltage. 

During the negative half -cycles represented by Fig. 5-3, the grid 
is negative; consequently, it does not attract any electrons from 
the plate-current stream crossing the tube. However, those 
electrons which had previously accumulated on the right-hand 
plate of C2 constitute the negative grid-bias voltage and will con- 
tinue to discharge downward through Rl. This action is made 
possible by the relationship between the large sizes of C2 and Rl 
and the time duration of an individual cycle of the signal current. 
The combination of C2 and R2 in this type of circuit will in- 
variably form a long time-constant network. 

C3 acts in conjunction with radio-frequency choke coil L4 as a 
filter whose purpose is to filter or remove the RF pulsations which 

94 



characterize the plate-current stream. As each individual pulsa- 
tion reaches the entrance to L4, it has a choice of flowing into the 
high impedance represented by the choke coil or the low im- 
pedance represented by the capacitor. Most of the strength of each 
pulsation will flow downward into C3 during the positive half- 
cycles of Fig. 5-2, driving an equal number of electrons downward 
from the lower plate of C3 into the ground connection. During the 
negative half -cycles depicted in Fig. 5-3, the plate current has 
diminished; therefore, the filter current through C3 reverses and 
flows upward. 

The action of the radio-frequency choke, when confronted by 
RF pulsations in plate current, is interesting. The purpose of any 
such choke is, of course, to prevent or inhibit the passage of high- 
frequency current. As each pulsation of plate current enters 
this choke coil at its left hand terminal, the increase in current 
which it represents generates a small current flowing in the 
opposite direction. This current has been shown in blue, and, in 



I. INPUT SIGNAL VOLTAGE 




2. GRID TANK VOLTAGE 
AFTER FEEDBACK HAS 
OCCURRED. 



3.GRID-LEAK BIAS 
VOLTAGE 



4. INDIVIDUAL PULSATIONS 
OF PLATE CURRENT 



5. AUDIO OUTPUT VOLTAGE 
MEASURED AT TOP OF L5- 




Fig. 5-4. Regenerative detector waveforms. 



Fig. 5-2, it flows from right to left, thereby reducing the total 
change in current which would otherwise occur, and also satisfy- 
ing the basic electrical property possessed by any inductance, 
namely, that it always reacts to any change in applied current 
flow in such a manner as to oppose that change. 

In Fig. 5-3, when the negative grid-driving voltage causes a 
reduction in the amount of plate current through the tube and 
eventually through coil L4, the choke coil again reacts in such a 
manner as to oppose this reduction in current. Thus, it brings into 
existence the small choke-coil current {shown in solid blue) , this 
time flowing in the same direction as the plate current, namely, 
from left to right. 

The actions of C3 and L4 are essentially independent actions. 
The capacitor action is intended to filter as much of the high- 
frequency component (the pulsations) of the plate current as 
possible to ground before it reaches the load, which is the primary 
winding of audio transformer L5. The choke-coil action is provided 
to prevent these pulsations from reaching the audio transformer. 
Thus, when the two components are put together they provide a 
highly effective filter combination, with a low-impedance path 
in the desired flow direction (to ground in this example) and a 
high-impedance path in the undesired flow direction (the output 
load and the power supply) . 

Fig. 5-4 shows a graphical representation of certain current 
and voltage waveforms at various points in the regenerative de- 
tector circuit. Line 1 of Fig. 5-4 shows the very weak nature of the 
input signal as it is received in LI. Line 2 shows essentially the 
same waveform, after it has been vastly amplified or regenerated 
by amplifier tube action plus feedback action. Line 2 represents 
the strength or amount of tank current which flows in the tuned 
circuit (L2 and CI). 

Line 3 (Fig. 5-4) indicates the relative strength of the grid-leak 
bias voltage which is detected by the grid-leak circuit combina- 
tion of C2 and Rl. Since any grid-leak bias voltage always consists 
of "trapped" or stored electrons, this voltage is always negative. 
We can see from a comparison of Line 3 with Line 2 that when a 
modulation peak occurs in the input signal, the individual RF 
cycles exhibit their maximum strength, and the leakage bias 
voltage reaches its maximum negative value. This is because the 
stronger RF cycles drive the grid to higher positive voltage values, 
and these higher positive voltages draw more grid leakage elec- 
trons out of the tube each cycle, and into storage on the right 
hand plate of C2. 

Line 4 of Fig. 5-4 indicates the complex nature of the plate 
current flowing through VI. Since tube current is always a one- 

96 



way current, or unidirectional in nature, the term pulsating DC is 
normally used to describe it. Inspection of Line 4 reveals that the 
plate current is pulsating at both a radio frequency and an audio 
frequency. Each cycle of the RF signal applied to the grid causes a 
small pulsation to occur in the plate-current stream. This action is 
accomplished primarily by the grid-driving current flowing up and 
down in Rl. Each cycle of audio-modulating voltage which is de- 
modulated by the grid-leak bias arrangement causes a larger (and 
slower) pulsation to occur in the plate current. 

Since L3 and L4 have extremely small amounts of inductance, 
the slow audio pulsations will pass through them without being 
affected. (The reactance of any inductor is directly proportional to 
the frequency of the current flowing through it.) 

Lines 3 and 4 point out that when the maximum negative grid- 
leak bias voltage exists on capacitor C2, the audio pulsation in the 
plate current stream is at a minimum. We can arbitrarily decide 
to label this period as a negative half -cycle of audio. The plate cur- 
rent flows downward at all times through the primary winding of 
the audio transformer. When plate current is approaching its 
minimum value, the steady decrease in plate current will induce 
a current flow downward in winding L6. This is the current which 
is shown in solid green in Figs. 5-2 and 5-3. 

When the plate current is approaching its maximum value, the 
steady increase in plate current will induce a current to flow up- 
ward in the secondary winding. Thus, a current which flows up 
and down at the basic modulation frequency through the head- 
phones is brought into existence. 

SUPERREGENERATIVE RECEIVER 

It is only a short and logical step from the regenerative detector 
circuit of Figs. 5-1, 5-2, and 5-3 to the superregenerative receiver 
shown in Figs. 5-5, 5-6, and 5-7. There are two important differ- 
ences between the two circuits: (1) the regenerative detector 
portion of the superregenerative receiver is permitted to oscil- 
late at the frequency of the signal being received, and (2) a 
separate circuit, called a quenching oscillator, is provided to stop 
the oscillations. 

Identification of Components 

The upper portion of Fig. 5-5 is a regenerative detector, similar 
in most respects (but not all) to the circuit explained previously. 
The components in Fig. 5-5 have been numbered wherever pos- 
sible to coincide with their counterparts in Fig. 5-1. The additional 
components which make up the quench oscillator are as follows: 

97 



R3 — Grid-leak biasing resistor, 

C5 — Tuned-tank capacitor. 

C6 — Grid-leak storage capacitor. 

C7 — Coupling capacitor. 

L7 — Coupling inductor. 

L8 — Tuned-tank inductor (autotransf ormer) . 

V2 — Quench-oscillator tube. 



DETECTOR -AMPLIFIER 

PLATE CURRENT 




I 






© 



© ® 



DETECTOR FEED- 
BACK 
CURRENT 




QUENCH OSCJ 

PLATE 
CURRENT 



OSCILLATOR 

FEEDBACK 
r CURRENT fCG, 




OSCILLATOR 

TANK 

CURRENT 



3 * r * 




=r(m 



GRID-LEAKAGE 

CURRENT 



GRID-DRIVING 
CURRENT 

Fig. 5-5. Operation of the superregenerative receiver — negative 
half-cycle of quench oscillator. 



98 



Identification of Currents 

The quenched-osc ilia tor portion of this circuit operates at a 
much lower frequency than the signal frequency being received. 
A quench frequency of about 20 kc is normal. The quenching os- 
cillator is a conventional Hartley oscillator. The various currents 
and the colors they are shown in are as follows: 




OSCILLATOR 

TANK 

CURRENT 



GRID-LEAKAGE 
CURRENT 



GRID-DRIVING 
CURRENT 

Fig. 5-6. Operation of the superregenerative receiver — positive 
half-cycle of quench oscillator. 



99 



1. Tank current (solid green). 

2. Grid-driving current (also in solid green). 

3. Grid-leak biasing current (dotted green). 

4. Plate current (solid red). 

5. Oscillator feedback current (dotted red). 

6. Autotransformer feedback current (also dotted red). 

7. Quenching current (solid red). 

Details of Operation 

Figs. 5-5 and 5-6 are chosen to represent two alternate half- 
cycles in the operation of the quenching oscillator. It must be 
recognized that because of the much higher frequency of the 
received signal (to which the regenerative detector is tuned), 
many cycles of operation of the detector portion of the circuit will 
occur during a single cycle of the quench oscillator. The re- 
generative detector is normally operated just below the point of 
oscillation so that a slight increase in the amount of inductive 
coupling between L2 and L3 will provide sufficient regeneration 
and feedback to cause the detector circuit to oscillate. 

This condition represents the maximum possible amount of gain, 
or amplification of the received signal. By itself, it is not an accept- 
able circuit technique, because once the oscillation is started it will 
continue even if the received signal goes off the air. Thus, the con- 
tinued oscillation would indicate a signal which is actually not 
present. To protect against this eventuality, the quench oscillator 
turns the oscillating detector circuit off at its own basic frequency, 
namely, 20,000 times each second. Once the oscillating detector 
has been stopped from oscillating, it cannot be started up again 
unless the desired input signal is again on the air. 

Fig. 5-5 depicts a positive half-cycle of the quench oscillator, 
when V2 is delivering a maximum surge of plate current. The 
oscillating tank current (shown in solid green) has moved down- 
ward through L3, thus delivering the electrons to the lower plate 
of C5, charging it to a negative voltage and making the upper 
plate positive. Whenever the upper plate is positive, it will draw 
the electrons of the grid-driving current upward through R3, 
making the upper end of R3 and the control grid of V2 positive. 
This releases a large pulsation of plate current into the tube. 

This plate current (shown in solid red) must first be drawn up- 
ward through the lower portion of L8. An increase of this plate 
current in the upward direction through the lower portion of 
L8 causes a separate feedback current to flow at an increasing rate 
in the downward direction through the entire winding. This feed- 
back current (shown in dotted red) flows in phase with the tank 
current, reinforcing it and supporting the quench oscillation. 

100 



The pulsation of plate current through V2 initially flows into 
the right-hand plate of C7, driving an equal number of electrons 
out of the left-hand plate and downward through L7 to ground. 
Because L7 is inductively coupled to L2, an upward flow of elec- 
trons will be induced in L2 during this half-cycle— this current 
is shown in dotted red. The polarity of the resulting "back electro- 
motive force" or back emf associated with this new current flow 
will be positive at the top of L2 and negative at the bottom. This 
positive polarity is also applied to the grid of VI, enabling it to 
conduct electrons from cathode to plate. Consequently, if a signal 
is being received through C4, oscillations will be set up in the 
tank circuit made up of L2 and CI. 

Fig. 5-6 depicts a negative half-cycle in the operation of the 
quench oscillator circuit. The electrons which make up the tank 
current in the quench oscillator have oscillated upward through 
L8, thereby delivering a large negative voltage on the upper plate 
of capacitor C5, as indicated by the green minus signs. Whenever 
this voltage is negative, it will drive electrons away from it along 
any available path. In this case, the only such path is downward 
through grid-driving resistor R3. (Electrons do not actually flow 
through C6 nor through any other capacitor for that matter, but 
normal capacitor action is such that when the negative voltage 
on C5 drives electrons onto the left-hand plate of C6, an equal 
number must flow out of the right-hand plate and downward 
through R3.) This downward movement of electrons through R3 
causes a negative voltage at the upper terminal of R3 and the grid 
of V2, causing an inevitable reduction in the plate current flowing 
through the tube. 

Two important actions stem from this reduction in plate current. 
First, the feedback current flowing in L8, (shown in dotted red) , 
reverses its direction of flow so that it again flows approximately 
in phase with the tank current, thereby replenishing or sustain- 
ing its oscillation. Second, the electron current which was pre- 
viously driven downward through L7 will now be drawn upward 
through this coil, and onto the left-hand plate of C7. This upward 
flow of current through L7 induces a counter current in L2, which 
is shown (in solid red in Fig. 5-6) flowing downward through the 
inductor. 

The back emf associated with this induced current in L2 is 
assumed to be negative at the top of L2 and positive at the bottom. 
A negative voltage at the top of L2 will cut off the electron flow 
through VI, and stop all of the current movements previously 
existing in the regenerative detector portion of the circuit. 

The input carrier signal may still be in existence and flowing in 
and out through C4; however, when it lacks the support of VI and 

101 



NORMAL AUDIO 

MODULATION 

ENVELOPE 



MANY RF CYCLES 



MODULATED CARRIER WHEN PERIODICALLY 
INTERRUPTED BY QUENCHING ACTION. 




Fig. 5-7. Modulated waveforms in quenching oscillator. 

the feedback arrangement between coils L3 and L2, it cannot 
sustain an oscillation in the tank circuit made up of L2 and CI. 

The quench circuit is provided against the specific contingency 
that the input carrier signal will periodically go off the air, and to 
stop, or quench, the oscillation going on in the RF tank circuit, L2 
and CI, just in case this contingency may have occurred. If it has 
not occurred, no harm will have been done, because on the next 
succeeding positive half -cycle of the quench oscillator (such as is 
depicted in Fig. 5-5), the oscillation of electrons in the RF tank 
will be "reignited," so to speak. If, however, the input carrier 
signal has gone off the air, the oscillation in the RF tank cannot 
again be started up during the next succeeding positive half -cycle 
of the quench oscillator. 

The ratio between the two oscillating frequencies will be per- 
haps 100 to 1, since the quench-oscillation frequency must be well 
above the range of the human ear — 20 kilocycles is a fairly stand- 
ard quench frequency — whereas the input carrier signal may well 
be 2,000 kc or higher. 

In the event a continuous carrier signal, such as speech or 
music, is being received, one might wonder what effect the 
quenching action would have upon the results. Fig. 5-7 is a 
simplified waveform diagram around which an explanation may be 
built. The first half of this illustration shows one audio cycle of a 
typical modulated carrier signal, and may be considered similar in 
all respects to the modulated waveforms in Lines 1 and 2 of Fig. 
5-4. The second half of Fig. 5-7 shows the modified waveform of 
the oscillation as it will actually exist in the radio-frequency tank 
circuit (L2 and CI) as a result of the quenching action. 

The signal and quenching frequencies have a ratio of 100 to 1. 
The RF oscillator circuit centered around VI will be turned on 



102 



for perhaps 50 cycles during a single positive half -cycle of the 
quench oscillator. During a negative half-cycle of the quench 
oscillator, the RF oscillation will be quenched for perhaps another 
50 cycles. If the audio frequency represented by the modulation 
envelope in the second part of Fig. 5-7 is 200 cps, 100 quenching 
operations will occur during a single audio cycle. These quench- 
ing operations will have no noticeable effect on the operation of 
the grid leak detector, or on the audio reaching the headphones, 
because of the time-constant relationship between C2 and Rl. 

When the values of C2 and Rl are multiplied together, the 
product is time in seconds. It is necessary only to choose the values 
of these two components so that their product in seconds will be 
long when compared to the time for one cycle of the quenching 
frequency (1/2 0,000th of a second) , and short when compared to 
the time for one cycle of the highest audio frequency expected 
to be received (1/5, 000th of a second) . When the proper choice of 
components has been made, it can be said that the grid leak bias 
voltage stored on C2 cannot discharge fast enough to follow the 
rise and fall in the audio modulation envelope. 

Thus, the grid voltage waveform shown in Line 3 of Fig. 5-4 
will be reproduced even while the quenching process is occurring 
and will be an accurate reproduction of the original audio modu- 
lation sent out from the transmitter. 



REVIEW QUESTIONS 



1. State the principal advantage and 
disadvantage of the regenerative 
detector circuit of this chapter. 

2. What is the function of the volt- 
age divider current which flows 
through R2 toward the power 
supply? 

3. In the regenerative detector of 
Fig. 5-1, describe the effect on 
circuit operation if the available 
tap on R2 is moved farther to the 
right. Also, if it is moved farther 
to the left. 

4. In the superregenerative re- 
ceiver of Fig. 5-5, what are the 



two approximate frequency 
ranges for the two oscillations? 
In this same circuit, what basic 
type of oscillator is used for the 
quench current? 

Describe the particular circuit 
action which couples the quench- 
ing action of (V2) to the opera- 
tion of the oscillating detector 
built around tube VI. 
If a continuous audio-modulated 
carrier is being received by the 
superregenerative detector, why 
will the quenching action not be 
noticeable to the listener? 



103 



Chapter 6 

TYPICAL SUPERHETERODYNE 
RECEIVER 



In this chapter the final two stages in a typical superheterodyne 
receiver — the IF amplifier and the audio power amplifier — will 
be discussed. Then the methods for checking voltages and mak- 
ing signal-substitution tests for the entire receiver will be pre- 
sented. Voltage checks and signal substitution tests are the two 
most common methods employed to isolate a trouble when servic- 
ing a receiver. 

IF AMPLIFIER 

The function of the IF amplifier stage is to increase the strength 
of (amplify) the signal voltage which is supplied by the mixer 
to the control grid of the tube to the level required by the 
detector. 

Identification of Components 

A typical IF amplifier circuit is depicted in Figs. 6-1 and 6-2. 
This circuit is composed of the following individual components, 
with functional titles as indicated: 

R12 — Cathode-biasing resistor. 

C7 — RF and IF filter or decoupling capacitor, 

Tl — IF input transformer (the secondary winding and the ca- 
pacitor in parallel with it are part of this circuit) . 

T2 — IF output transformer (the primary winding and the ca- 
pacitor in parallel with it, are part of this circuit) . 

V2 — Pentode tube used as IF amplifier. 

104 



Identification of Currents 

The following electron currents are at work in this amplifier 
circuit: 

1. Grid -tank current (dotted blue) . 

2. Plate current (solid red) . 

3. Screen -grid current (dotted red) . 

4. Plate-tank current (solid blue) . 

5. IF filter current (dotted red). 

Details of Operation 

The oscillation of electrons (shown in dotted blue) which oc- 
curs in the grid-tank circuit drives the control grid of V2 to 
alternate positive and negative values. It is supported, or re- 
plenished, each half-cycle by the movements of the tank current 
which flows up and down through the primary winding of 
Tl. Fig. 6-1 shows the current in the primary winding flowing 
downward and the current in the secondary winding flowing 
upward. This action delivers electrons to the top of the tank, 
where they become concentrated on the upper plate of the tank 
capacitor, constituting a negative voltage at this point. Conse- 
quently, this half-cycle has been labeled as a negative half-cycle 
of operation. 

Fig. 6-2 shows the current in the primary winding moving 
upward and the tank current moving downward through sec- 
ondary winding of Tl. This action delivers electrons to the lower 
plate of the tank capacitor, creating a deficiency of electrons on 
the upper plate. This deficiency constitutes a positive voltage; 
therefore Fig. 6-2 has been labeled as a positive half-cycle of 
operation. 

Plate Current — During negative half-cycles (Fig. 6-1), the 
control grid of V2 will have its most negative voltage, and the 
plate current electron stream will be reduced to its minimum 
value. During the positive half-cycles (Fig. 6-2) , the plate cur- 
rent stream is increased to its maximum value. Thus, the plate 
current is a form of pulsating DC, which flows continuously from 
cathode to plate within the tube, then downward through the 
primary winding of T2 to the power supply. 

The pulsations in this plate current coming out of the tube 
support a new oscillation of electrons in the plate-tank circuit. 
This electron current is shown in solid blue. In Fig. 6-2 when 
another pulsation of electrons arrives from the tube, it reaches 
the upper plate of the tank capacitor simultaneously with the tank 
current which has moved upward through the primary winding 

105 



AMPLIFIED 

IF TANK 

CURRENT 

(SECONDARY) 




AMPLIFIED 

IF TANK 
CURRENT 
(PRIMARY) 

*" TO POWER 
SUPPLY 



FILTERING 
OR DECOUPLING 
CURRENTS 



PLATE AND 

SCREEN-GRID 

CURRENT 

FROM VI 

TO AVC -* 

STORAGE 

CAPACITOR 



Fig. 6-1. Operation of an IF-amplifier circuit — negative half-cycle. 

of T2, thereby reinforcing this tank current. The plate-tank cur- 
rent is an amplified version of the signal current which flows in 
the grid-tank circuit. This means that it is a stronger oscillation, 
or in other words that a larger quantity of electrons is oscillating 
in the plate tank than in the grid tank. 

Another current (also shown in solid blue) is shown flowing in 
the secondary winding of T2 in Fig. 6-2. It is sustained by the 
primary tank current which oscillates up and down through the 
primary winding. As the primary current moves upward (Fig. 
6-2) , the secondary current moves downward; as the primary 
current moves downward (Fig. 6-1) ( the secondary current moves 
upward. 

Screen-Grid Currents — The screen-grid current, and its asso- 
ciated filter current (both shown in dotted red) are the final set 
of electron currents in this circuit. The screen-grid current con- 
sists of electrons which are captured from the plate current 
stream within the vacuum tube. The screen grid has a fairly 
high positive voltage on it so that some of the negative electrons 
of the plate-current stream adhere to the wires of the screen 
grid. These electrons exit from the tube and rejoin the plate cur- 
rent below T2 and continue on to the power supply. 



106 



PLATE 
CURRENT 



PRIMARY SECONDARY 
IF TANK y~^ IF TANK 
CURRENT (Ty CURRENT 



\ 



SCREEN- GRID 
CURRENT 

L 



AMPLIFIED 
IF TANK 
CURRENT 
(SECONDARY) 







PLATE AND 

SCREEN-GRID 

CURRENT 

FROM VI 

TO AVC « 5 * 

STORAGE 

CAPACITOR 



AMPLIFIED 

IF TANK 
CURRENT 
(PRIMARY) 

TO POWER 
SUPPLY 



IF FILTERING 
OR DECOUPLING 

CURRENTS 



Fig. 6-2. Operation of an IF-amplifier circuit — positive half-cycle. 

Like the plate current, the screen-grid current is also a pulsating 
DC. The pulsations occur during the positive half -cycles shown 
in Fig. 6-2. In Fig. 6-1 when the control grid is negative, the elec- 
tron stream within the tube is reduced to its minimum value, 
and the plate and screen-grid currents are also reduced. 

C7 filters out the fluctuations, or pulsations, in the plate and 
screen-grid currents. Fig. 6-2 shows such a filter action occurring. 
When a pulsation occurs in the screen-grid current, electrons 
flow onto the upper plate of C7, and other electrons are driven 
away from the lower plate to ground. 

In Fig. 6-1, when no such pulsation occurs, electrons flow off 
the upper plate of C7 and into the power supply. This permits 
other electrons to be withdrawn from ground, and flow onto the 
lower plate of C7. The action of filtering out the pulsations in tube 
currents before they reach the power supply is frequently re- 
ferred to as decoupling of the power supply. 

AVC Voltage — The bottom of the grid-tank circuit is connected 
directly to the AVC storage capacitor (explained in Chapter 2) . 
Since this capacitor has a permanent negative voltage stored 
on it, this voltage is also applied to the grid of V2 and acts as 
a permanent biasing voltage on that tube. Thus, as the AVC 



107 



voltage varies, the gain of V2 is changed. The variations in grid 
voltage caused by the oscillation of electrons in the grid tank 
of this tube will alternately add to or subtract from this per- 
manent biasing voltage. 

AUDIO POWER AMPLIFIER 

The basic function of an audio power amplifier is to increase 
the strength of the audio signal delivered to it by the preceding 
amplifier, and to generate a heavy electron current at these 
same audio frequencies in order to operate the speaker. 

Identification of Components 

A typical audio power-amplifier circuit is illustrated in Figs. 
6-3 and 6-4. This circuit is composed of the following individual 
components j with the functions indicated: 

R8 — Grid-driving resistor. 

R9 — Cathode-biasing resistor. 

RIO — B+ dropping resistor (actually part of the power-supply 

filter system) . 
C12 — Coupling and blocking capacitor. 
C13 — IF filter capacitor. 
T3 — Audio output transformer. 
V4 — Pentode power-amplifier tube. 
SP1 — Speaker. 

Identification of Currents 

The following electron currents are at work in this power- 
amplifier circuit: 

1. Grid-driving current (dotted green) . 

2. Plate current (solid red) . 

3. Screen-grid current (dotted red) . 

4. Speaker current (solid blue) . 

5. Plate and screen-grid currents from other tubes (also in 
solid red) . 

Details of Operation 

The grid-driving current (shown in dotted green) moves up 
and down through R8 at the audio frequencies being amplified. 
This current is in turn driven by the pulsations in plate current 
from the previous tube. Fig. 6-3 shows one such pulsation occur- 
ring, with plate current (shown in solid red) flowing onto the 
left-hand plate of C12. This action drives an equal number of 

108 



electrons away from the right-hand plate of C12 and downward 
through grid resistor R8. 

The grid-driving current flowing through R8 causes the volt- 
age at the top of R8 to be negative. For this reason, Fig. 6-3 has 
been labeled as a negative half-cycle of operation. In Fig. 6-4, 
when the plate current pulsation is flowing out of the left-hand 
plate of C12, it draws the grid-driving current upward through R8. 
This indicates that the top of the resistor is more positive than 
the bottom. 

In Fig. 6-3 when the top of R8 has its most negative voltage 
value, the plate-current electron stream flowing through V4 will 
be reduced to its minimum value. In Fig. 6-4 when the grid is 
positive, maximum plate current will flow. 

Poiuer Amplification — The construction of a power amplifier 
tube differs somewhat from that of a voltage-amplifier tube. A 
power-amplifier tube is constructed so that it will conduct a much 
heavier plate current at full conduction than a voltage-amplifier 
tube delivers. Electrical power varies as to the square of the 
current; therefore an amplifier tube which is capable of delivering 
wide extremes of electron current has been given the name of 
power amplifier. This title can be misleading because all amplifier 
tubes deliver their plate currents into some kind of a load, and 
consequently, some power is developed in each of these loads 
by these currents. A tube is called a power amplifier when it 
delivers a heavy enough current into its load to develop an ap- 
preciable amount of power. 

The basic formula for computing power across a resistive 
load is: 

F = PR 
where, 

P is the power developed in watts, 

I is the current flowing through the resistor in amperes, 

R is the resistance of the load in ohms. 

The formula for computing power developed across an induc- 
tive load, such as the primary winding of T3, is: 

P = PX 
where, 

P is the power in watts, 

I is the current through the inductor in amperes, 
X is the inductive reactance of the transformer primary in 
ohms. 

Plate Current — The plate-current path starts at ground below 
R9. The electrons which make up this current flow upward 

109 



PLATE CURRENT 
PLATE CURRENT (MINIMUM) 
PULSATION FROM V3 



PERMANENT 
MAGNET 




sC~ 



ONE HALF-CYCLE 

OF SOUND WAVE CAUSED 

BY COMPRESSION OF AIR 



PLATE AND 

SCREEN CURRENTS 
FROM OTHER TUBES 



TO CATHODE 
OF RECTIFIER 



Pig. 6-3. Operation of an audio-output circuit — negative half -cycle. 

through R9, through the vacuum tube from cathode to plate, 
downward through the primary winding of T3, from where it 
flows directly to the cathode of the rectifier tube. All of the other 
plate and screen-grid currents join with this plate current at the 
right-hand end of RIO, to be drawn eventually through the recti- 
fier tube and delivered back to ground. 

Output Transformer — Transformer action can be a very difficult 
physical action to visualize. Stated in the simplest terms, when 
a current, such as the plate current of V4, can be made to pulsate 
through one winding of the transformer, it will cause another 
current to flow back and forth in the other winding. The pulsa- 
tions of plate current flowing downward through the primary 
winding of T3 cause the speaker current (shown in solid blue) 
to flow in the secondary winding. 

When this plate current is increasing, as it does during the 
positive half-cycle of Fig. 6-4, the speaker current increases in 
the upward direction through the secondary winding. When the 
plate current decreases during the negative half -cycles (Fig. 6-3) , 
the speaker current increases in the downward direction through 
the secondary winding. Since the plate current consists of con- 
tinuous pulsations at the various audio frequencies, the speaker 
current will flow up and down through the secondary winding 



110 



PLATE CURRENT 
PULSATION FROM V3 



PLATE CURRENT 

(MAXIMUM) 



PERMANENT 
MAGNET 




grid-driving(R9) > 

CURRENT 1 



PLATE AND 
SCREEN CURRENTS 
FROM OTHER TUBES 



ONE HALF-CYCLE 

OF SOUND WAVE CAUSED 

BY RAREFACTION OF AIR 



TO CATHODE 
OF RECTIFIER 



Fig. 6-4. Operation of an audio-output circuit — positive half-cycle. 

(and through the moving coil of the speaker) at these same audio 
frequencies. 

Like all similar output transformers, T3 has more turns in 
its primary winding than its secondary winding. This means that 
it is a current step-up transformer. (A current step-up trans- 
former is the same thing as a voltage step-down transformer.) 

The determination as to whether a transformer will step the 
current up or down is governed by a simple formula, which 
states: 

LN. 



I«N P 
where, 

I y is the current flowing through the primary winding, 
I s is the current flowing through the secondary winding, 
N„ is the number of turns of wire in the secondary winding, 
N,, is the number of turns of wire in the primary winding. 

Speaker Action — The speaker is a typical moving-coil arrange- 
ment, which involves the use of two magnets — a permanent mag- 
net and a temporary magnet (formed by the speaker voice-coil 
winding) . The permanent magnet has been shown with its 
south pole adjacent to the left-hand end of the temporary magnet. 



Ill 



During the negative half-cycles shown in Fig. 6-3, the speaker 
current flows through the speaker voice coil in such a direction 
as to make the left end of the temporary magnet have a south 
magnetic pole. This causes it to be repelled by the adjacent south 
pole of the permanent magnet, and it moves to the right. Since 
the speaker cone is connected to the voice coil, it also moves. The 
movement of the speaker cone compresses the air in front of 
it and causes a half-cycle of a sound wave. 

During the positive half -cycles depicted by Fig. 6-4, the speaker 
current flows in such a direction as to create a north magnetic 
pole at the left end of temporary magnet. This north magnetic 
pole will be attracted by the south magnetic pole of the per- 
manent magnet, and its movement will cause a rarefaction of the 
air in front of the speaker diaphragm. This rarefaction of the 
air constitutes a second half-cycle of the sound wave. 

Lenz's Law — The rule for relating the direction of electron 
flow around an iron core to the resulting magnetic polarity of that 
core is known as Lenz's law. If the iron core is grasped by the 
left hand so that the fingers point in the direction that the electron 
current is flowing through the coil, then the thumb points toward 
the temporary north magnetic pole of the electromagnet. 

A moving-coil speaker is a current-operated device. The amount 
of movement of the iron core and the diaphragm to which it is 
attached depend on the amount of current flowing through the 
moving coil each half -cycle. The amount of diaphragm movement 
determines the loudness of the sound coming from your radio. 
Output transformers, such as T3, have more turns in their pri- 
mary winding than in their secondary winding so that large 
fluctuations in plate current will be increased even more by the 
current step-up relationship between the windings. 

VOLTAGE CHECKING THE SUPERHET RADIO 

When the superhet radio becomes inoperative, or "dead," there 
are a number of tests, progressing from very simple ones to more 
complex ones, which the technician can use to locate and remedy 
the source of trouble. Probably the first of these tests is a visual 
check to see if all the tubes are lighted. Some tubes may have 
metal envelopes, but in simple table model radios, most if not 
all tubes will have glass envelopes. When the radio is turned on, 
we can look inside the envelope (after approximately 18 seconds 
have elapsed) and see a red glow that indicates that the cathode 
is heating normally. In tubes with metal cases, we can usually 
tell whether the filament is heating by feeling the metal envelope 
to see if it is warm. 

112 



Fig. 6-5 shows the circuit diagram of a typical radio, with the 
filament heating current (in solid blue) flowing through its in- 
tended path. (Partial schematics of the various stages in this 
radio have been presented in the previous chapters.) The fila- 
ments for all five tubes are connected in series, so the same 
heating current must flow through every filament. If a single 
tube filament fails, no current can flow in any of them, so each 
must be checked on a tube tester. When the faulty tube is located 
in this manner, and replaced, the radio will usually operate again. 

Once it is determined that the tubes are heating normally, the 
simplest method of isolating a faulty component is by the process 
known as voltage checking. This process requires only a single 
piece of test equipment — a standard voltmeter, along with a 
diagram showing voltages which can be expected at each tube 
electrode during normal operation. Each and every one of these 
electrode voltages has a value which is determined by one or 
more electron currents flowing through or along a certain re- 
sistive path. When each such current is understood and visualized, 
and when its complete path is recognized, the student or tech- 
nician can infer a great deal about that current and about the 
components through which it flows by noting the voltage at the 
electrode where each current enters or exits from the tube. 

The plate currents are superimposed on the circuit diagram 
in Fig. 6-5 in solid red. Each of these currents is drawn up from 
ground below the respective cathodes of the tubes, and across 
the tube by the high positive voltage at the plate. The point of 
highest positive DC voltage in the radio is the cathode of the 
half -wave rectifier tube V5; this positive voltage draws all of 
the plate currents through their respective plate-load circuits 
(primary of Tl, primary of T2, R7, and primary of T3) to the 
filter circuit (RIO, C14, and C15) and to the cathode of V5. 

The DC voltages which exist at the principal electrodes of the 
vacuum tubes are all directly associated with the flow of these 
currents. A great deal of servicing information can be inferred by 
observing these DC voltages. 

Voltage Checking VI 

The schematic in Fig. 6-5 tells us that the cathode of VI should 
be at ground voltage, the second control grid at —12 volts, and 
the screen grid and plate at +105 volts. 

The negative or common ground terminal of the voltmeter 
should be applied to the chassis of the radio, and the positive 
probe should be touched in turn to each of the electrodes. 

It could happen that L2 in the oscillator tank has burned out 
and opened so that the flow of plate current enroute from ground 

113 



to the cathode is interrupted. When this happens, the actual volt- 
age existing at the cathode will be slightly positive. Even though 
no plate current can flow if the ground connection is broken, some 
electrons will be drawn across the tube by the high positive plate 
voltage, creating an electron deficiency on the upper plates of 
C4 and C5. However, this small positive voltage cannot be meas- 
ured unless an electrostatic voltmeter is employed. 

When the normal test voltmeter is used, a much higher than 
normal positive voltage will be measured at the cathode ter- 
minal when L2 is open (it might be 75 to 100 volts in this in- 
stance) . The reason for the high positive voltage is that the 
instant the meter probe touches the cathode terminal, the in- 
ternal resistance of the meter completes the cathode circuit to 
ground. The meter resistance will be very high compared to the 
internal plate resistance of the tube. These two resistances form 
a voltage divider across the power supply. Therefore, even though 
the plate current is greatly reduced, a small current will flow 
through the meter and the tube. Because the meter resistance 
is the larger of the two resistances forming the voltage divider, 
most of the power supply voltage will be dropped across the 
meter resistance and read on the meter scale. 

Note: Only an open-circuit in the lower half of L2 will cause 
this trouble. An open circuit in the upper half of L2 will not 
interrupt the flow of plate current, although it will surely stop 
the flow of oscillator tank current (shown in dotted green) . This 
defect can be detected by signal substitution followed by a con- 
tinuity check of the inductor. 

If the measured cathode voltage proves to be zero, this does 
not necessarily prove that plate current is flowing, since the 
cathode voltage will also be zero when no plate current flows. 
The voltmeter probe should now be moved to the screen grid of 
VI, and then to the plate. Each of these electrodes should indi- 
cate approximately 105 volts, or only slightly less than the B+ 
voltage of 110 V. Both of these tube currents must flow through 
R2 enroute to the power supply, and the voltage drop across R2 
accounts for the 5-volt difference between 105 and 110 volts. 

If R2 has opened, these tube currents cannot flow; therefore, 
the voltage at the plate and the screen grid will be close to zero. 
An internal short in C7 would also result in this same symptom 
of zero voltage at the plate and screen grid. Heavy current would 
be flowing up from ground through C7, and to the right through 
R10 to the power supply. 

The simplest check to determine which of these two compo- 
nents may have failed is to bridge R2 with another resistor of 
approximately equal value. If R2 has failed, this procedure will 



114 




@® ® 



ismuz 




i-^wh 



Fig. 6-5. Operation of a typical radio receiver. 



115 



provide an alternate path for the tube currents from VI, and the 
set may begin operating normally. If the capacitor has failed 
due to an internal short-circuit, one of its terminals must be 
unsoldered before the standard capacitor check can be made. 

The AVC voltage can also be checked at this time, since it 
is the —12 volts existing at pin 7 of the tube. The AVC voltage 
is the accumulation of electrons which are stored on the left- 
hand plate of CI. These electrons are originally part of the de- 
tector current which flows through the diode portion of V3 and 
eventually finds its way through the AVC resistor R3 to CI. If 
CI should become shorted, these electrons cannot accumulate 
but will flow or leak through CI to the common ground connec- 
tion between CI, C2, and C3. 

A short circuit in either C2 or C3 would have the same effect 
of diverting all of the stored AVC electrons to ground, and of 
holding the second control grid of the tube at ground voltage. 
Each of these three capacitors should be given the standard check- 
out when this symptom occurs. 

Voltage Checking V2 

The positive probe of the voltmeter can be applied in succes- 
sion to the cathode, screen grid, and plate of V2. This gives valu- 
able information about the components in the circuit. The posi- 
tive cathode voltage of .86 volts exists solely because of the 
upward flow of plate current through R12. If the cathode voltage 
is significantly higher than this value, it probably means that 
the resistor has opened or increased in value. If R12 is open, a 
high positive voltage will be measured on the cathode as explained 
for VI. Also, the plate and screen voltages on V2 (and the other 
tubes) will increase by 5 to 10 volts. The reason for the increase 
of plate and screen voltages is that, with R12 open, V2 plate and 
screen current no longer flow through power supply filter resistor 
RIO and the voltage drop contributed by these currents no longer 
appears across this resistor. Temporarily bridging R12 with a low 
value resistor or even a straight wire should restore the set to 
operation and prove that R12 needs to be replaced. 

Zero volts at the cathode indicates that R12 is probably all 
right, but tube currents are not flowing through it. A value be- 
tween zero and +.86 volts at the cathode would mean that the 
tube currents are flowing in reduced quantity — an indication of 
reduced emission of the tube. This fact can be verified on a tube 
tester. 

The plate and screen-grid voltages should also be measured. 
A value of 110 volts indicates that the tube currents are flowing 
normally. Any higher value indicates that these currents (or those 

116 



of some other tube) are flowing in reduced amounts, or not at all. 
If the plate and screen-grid voltages are zero, this would mean 
that the line to the power supply is probably open. 

If the screen-grid voltage is normal, and the plate voltage is 
zero, the primary winding of T2 could be open, so that screen 
current can flow, but plate current cannot. 

The AVC voltage of -12 volts can be measured at the control 
grid of V2, just as it was measured at the second control grid 
of VI. 

Voltage Checking V3 

The AVC voltage of -12 volts will exist at the two control 
grids previously mentioned, as well as at the two diode plates of 
V3. The mere existence of this voltage tells us that the tuned 
circuit associated with T2 is functioning normally and that the 
electron current through the diode portion of V3 is flowing. 

A positive value of 50 volts should exist at the plate of V3. 
The difference between this value and the 110 volts of the B+ 
line (a 60-volt difference) is the voltage drop across the 470,000- 
ohm resistor, R7, caused by the flow of triode plate current 
through R7. If we find zero volts at the plate, the flow of plate 
current has probably been interrupted by R7 opening, or Cll 
may be shorted. 

If the full value of 110 volts exists at the plate, we know that 
R7 is all right, but that no voltage drop exists across it because 
no plate current flows through it. This means that the tube is not 
conducting. Such a condition could be caused by a faulty tube, 
or the cathode connection to ground may have opened. 

Voltage Checking V4 

If the electrode voltages on the three previous tubes have 
checked out satisfactorily, the cathode, screen grid, and plate 
of V4 should be checked. The desired cathode voltage should be 
+7.5 volts. This is the amount of voltage drop across R9 caused 
by the upward flow of plate current through it. If for any reason 
this current cannot flow, the cathode voltage will be at ground 
voltage, as long as the current path through R9 has not been 
interrupted. If R9 has opened, however, the cathode will be at 
a high positive voltage. 

The plate should be at 120 volts, since it is connected directly 
to the point of highest voltage in the radio, namely, the cathode 
of rectifier tube V5. The plate current must flow through only 
the low resistance of the primary winding of T3. If the V4 plate 
voltage is zero, it most likely means that the primary winding of 
T3 is open. 

117 



If the plate voltage of V4 is equal to the full power-supply 
voltage of 125 volts, it means there is no voltage drop across the 
primary winding of T3 because no plate current is flowing 
through V4. 

Zero voltage at the screen grid could mean that either C14 
or C7 has shorted to ground. Either of these two defects would 
also have placed zero voltages at the plates and screen grids 
of VI, V2, and V3, and should have been detected in the earlier 
tests. 



SIGNAL SUBSTITUTION 

After the electrode voltages have been checked and found 
satisfactory, a somewhat more complicated procedure known as 
signal substitution becomes necessary, in order to locate which 
component or components may have failed. To perform this 
series of tests, a standard signal generator is required. The signal 
generator must be able to provide: 

1. An unmodulated RF signal at the highest local oscillator 
frequency to be encountered — approximately 2,000 kc. 

2. A modulated carrier signal which can be varied from the 
IF to the top of the broadcast band — 455 kc to 1,600 kc. 

3. A 400-cps audio note. 

The principle of signal substitution is to apply an artificially 
generated signal of the type normally encountered at each of 
several significant points within the receiver and to observe 
the operation of a limited portion of the receiver under these 
controlled conditions. The significant points in this case are 
principally the control grids and the plate circuits of each of the 
amplifier tubes. The normal signal one expects to find at each of 
these points is as follows: 

1. Antenna input coil (LI)— a very weak modulated RF signal. 

2. Second control grid of VI — same very weak modulated RF 
carrier signal. 

3. First control grid of VI — a weak oscillatory signal whose fre- 
quency must always be 455-kc higher than the incoming 
carrier signal. 

4. Plate circuit of VI— a weak RF signal at intermediate fre- 
quency of 455 kc. 

5. Control grid of V2 — a weak but amplified version of the 
IF signal at the plate circuit of VI. 

6. Plate circuit of V2 — an amplified version of the IF signal 
found in the grid circuit of V2. 

118 



7. Diode detector tank circuit— an IF signal of approximately 
equal strength to that in 6. 

8. Junction of R4 and R5 — a weak audio voltage. 

9. Control grid of triode portion of V3 — the same weak audio 
signal. 

10. Plate of V3 — an amplified version of this same audio signal. 

11. Grid circuit of V4 — same audio signal as in 10. 

12. Plate circuit of V4 — amplified version of same audio signal. 

13. Speaker coil — same audio signal. 

As with all other measurement devices, the signal generator 
method requires some kind of an output indicator. The speaker 
of the receiver serves this purpose very well. The procedure for 
checking out a receiver by the signal substitution method is to 
start at the bottom of the foregoing list and apply the appropriate 
signal to each point. If that portion of the circuit is functioning 
properly and if all prior portions of the circuit have been pre- 
viously checked out (working backwards from the bottom of 
the list, of course) , then a 400 -cycle note should be heard from 
the speaker. 

The reason the first check is made at the speaker is to be cer- 
tain that our so-called "output indicator" is functioning properly. 
You must work backwards in an orderly fashion from plate to 
grid because when the faulty circuit is reached, there can be no 
question as to which circuit is at fault. We can only be sure of 
this if we know that all intervening circuits from the point in 
question to the output circuit (speaker) are functioning properly. 

With the common ground of the signal generator connected to 
the common ground of the radio, the output probe of the genera- 
tor should be touched to the top of T3. In order for the speaker 
to function properly, a closed circuit must exist through the 
secondary winding and through the speaker winding. The 
audio current (shown in green) must flow back and forth through 
this path. The 400-cycle setting of the signal generator pro- 
vides this current, but a closed path through these two wind- 
ings must exist in order for a 400-cycle note to be heard from 
the speaker. 

Once the operability of the speaker has been established, the 
signal generator probe should be moved to the plate of V4. In 
normal operation, the plate current for this tube, shown in red, 
will pulsate at some audio frequency through the tube and down- 
ward through the primary winding of T3. The signal generator 
probe provides these pulsations for us, and if transformer T3 
is functioning properly, an audio current will be induced in the 
secondary winding, and operate \he speaker. If no signal is heard, 

119 



then either the plate circuit is grounded, or T3 is not functioning. 
A grounded plate circuit would have been discovered on the prior 
voltage check. 

If the signal is heard on this test, the signal-generator output 
should be turned down slightly, to allow for the gain which V4 
should provide, and the probe should be applied to the control 
grid of V4. The probe will provide the grid driving current which 
has been shown in solid green in Fig. 6-5. This current moves up 
and down through R8, and develops the alternating positive and 
negative voltages which "drive" the control grid. If the tube and 
its associated circuitry are working properly, another 400-cycle 
note should be heard from the speaker. 

With the generator still providing the 400-cycle note, the probe 
should be applied in succession to the plate and then the control 
grid of the triode portion of V3. Whenever the probe is moved 
to include an additional amplifying device such as a tube, care 
should be taken to reduce the output level of the generator to 
keep the speaker volume at a reasonable level. 

The normal signal expected at the diode plates of V3 is an IF 
voltage which carries some audio modulation. Therefore, the 
signal generator should be reset to provide this kind of output, 
and the probe can be touched to either of the diode plates. The 
signal generator will then stimulate the appropriate IF tank cur- 
rent (shown in solid green in Fig. 6-5) in the tank circuit. Each 
positive half-cycle will cause one shot or pulsation of electrons to 
flow from the cathode to the diode plates of V3. 

If resistors R4 and R5 and capacitors C8 and C9 are functioning, 
the tuned-tank circuit is all right and the detector portion of V3 
is conducting normally, the diode current will flow as shown in 
solid green. Finally, if CIO and R6 are functioning, an audio fre- 
quency current will be driven back and forth through CIO, and 
up and down through R6. Since we have already checked out all 
following circuits, this audio current flow through R6 should 
"drive" the amplifier portion of V3 and cause the 400-cycle note 
to be heard in the speaker. 

When the workability of this much of the circuit has been 
established, the generator probe should be moved to the earlier 
check points in reverse order to that appearing in the list of check 
points. The next three check points all require the same type of 
signal during normal operation, namely, the IF signal modulated 
with 400-cycle audio. Since our generator is already set to provide 
this output, the probe should be applied in turn to the plate and 
control grid of V2, and then the plate of VI. The purpose in each 
case is to supply the normal signal current, which is shown in 
solid green in the four IF tank circuits of Fig. 6-5. 

120 



After these points have been checked, the signal generator 
should be reset to provide a substitute signal for the local oscilla- 
tor. A value of 455 kc should be added to the frequency indicated 
on the tuning dial of the radio to determine the generator setting. 
For instance, if the receiver is turned to a known station which 
radiates on a frequency of 1,000 kc, the generator should be set 
to provide an unmodulated signal at 1,455 kc. By rocking the 
tuning dial slowly back and forth, it may be possible to "bring in" 
the desired station. 

If none of the preceding tests has helped locate a faulty circuit 
component, a modulated RF signal from the generator should 
be applied to the top of the antenna coil (primary winding LI) . 
This winding may have opened, either at the ground connection 
or at the point where it connects to the antenna. An open circuit 
at either spot would effectively prevent the necessary flow of the 
antenna current through this winding, and can account for an 
inoperative receiver. 



REVIEW QUESTIONS 



In the IF amplifier circuits of 
Figs. 6-1 and 6-2, does the plate 
current of tube V2 flow intermit- 
tently or continuously? Is it a 
pulsating direct current, a pure 
direct current, or an alternating 
current? 

In these same figures, four sepa- 
rate oscillating currents are 
shown in blue in four separate 
tank circuits. At what frequen- 
cies are these oscillations occur- 
ring? 

The plate and screen grid cur- 
rents must flow through R12 to 



reach tube V2, developing a 
voltage across R12. What is the 
polarity of this voltage and what 
functional name is usually as- 
signed to it? 

In what fundamental respect 
would the plate current of the 
power amplifier tube V4 in Figs. 
6-3 and 6-4 differ from the plate 
current of an audio voltage am- 
plifier operating at the same fre- 
quency? 

Is audio output transformer 
T3 a voltage step-up or step-down 
transformer? 



121 



Chapter 7 

TYPICAL TRANSISTOR 
RECEIVER 

In this chapter a typical transistor broadcast receiver will be 
analyzed. First, the operation of the receiver and each electron 
current flow will be explained. Then the methods of checking the 
various voltages will be outlined. Finally, signal substitution tests 
will be given. 

TRANSISTOR BROADCAST RECEIVER 

Figs. 7-1, 7-2, and 7-3 show three identical circuit diagrams of 
a typical transistor broadcast receiver. All of the electron currents 
which flow in this receiver during normal operation have been 
shown and identified in Fig. 7-1. In Figs. 7-2 and 7-3 these cur- 
rents have been separated so that their actions may be more 
easily analyzed. 

This receiver utilizes the superheterodyne circuit. Two stages 
of IF amplication are employed followed by a solid-state diode 
detector. The audio signal developed at the detector is then 
amplified by two voltage-amplifier stages, and push-pull power- 
amplifier stage. An adidtional stage is provided for overload 
protection. 

Identification of Components 

The individual components which make up this radio and their 
functions are as follows: 

Rl — Volume control. 

R2, R3 — Base-biasing resistors for XI. 

122 



R4 — Emitter-biasing resistor for XI. 

R5 — Oscillator-tank damping resistor. 

R6 — Emitter-biasing resistor for X3. 

R7, R8 — Base-biasing resistors for X2. 

R9 — Emitter-biasing resistor for X2. 

RIO — Base-biasing resistor for X3. 

Rll — Collector resistor for X3. 

R12, R13 — Base-biasing resistors for X4. 

R14 — Emitter-biasing resistor for X4. 

R15, R16, — Base-biasing resistors for X5. 

R17_Collector-load resistor for X5 and base-biasing resistor 

for X6. 
R18 — Emitter-biasing resistor for X6. 
R19 — Power-supply decoupling resistor. 
R20 — Emitter-biasing resistor for X7 and X8. 
R21 — Base-biasing resistor for X7. 
R22 — Base-biasing resistor for X8. 
CI — AVC capacitor. 
C2A — Emitter bypass capacitor for X6. 
C2B — Power-supply decoupling capacitor. 
C3 — Input coupling and blocking capacitor. 
C4 — Oscillator-tank coupling capacitor. 
C5 — Bypass capacitor for R6, 
C6 — Feedback and neutralizing capacitor for X2. 
C7 — Emitter bypass capacitor for X2. 
C8 — Bypass capacitor for RIO. 

C9 — Bypass capacitor for R12. 

CIO — Feedback and neutralizing capacitor for X4. 

C12 — IF filter capacitor. 

C13 — Audio coupling and blocking capacitor. 

C14 — Parasitic suppression capacitor. 

C15, C16 — RF tank capacitors. 

C17, C18 — Oscillator tank capacitors. 

LI — Antenna-coupling transformer. 

L2 — Oscillator-tank transformer. 

L.3 — First IF transformer. 

L4 — Second IF transformer. 

L5 — Third IF transformer. 

Tl — Audio interstage transformer. 

SP1— Speaker. 

XI — 2N412 Converter transistor. 

X2, X4— 2N410 IF amplifier transistors. 

X3 — 3458 Overload transistor. 

X5, X6 — 2N406 Audio-amplifier transistors. 

X7, X8 — 2N408 Audio output transistors. 

123 



Ml — Battery power supply. 
M2 — Diode detector. 

Note that all of the foregoing transistors are PNP type. It is 
standard practice to use one type of transistor — either PNP or 
NPN — in small systems, such as this one. This practice provides 
simplification in biasing and power-supply requirements. 

Identification of Currents 

There are approximately 50 different electron currents at work 
in this radio during normal operation. These currents may be sub- 
divided into seven main types, as follows: 

1. Base-emitter current within each transistor — there is one 
such current for each transistor, making eight altogether 
(solid green) . 

2. Collector-emitter current within each transistor — there is 
one such current for each transistor, making eight in all 
(solid red). 

3. Voltage-divider currents — there are six voltage-divider cur- 
rents (solid blue). All of them originate at a ground con- 
nection and are drawn through certain resistors to the 
positive terminal of the power supply. These resistor net- 
works, and the important function provided by the current 
through each one, are: 

First Voltage-Divider Current (R2, R3, and R19)— Pro- 
vides a small positive base-biasing voltage for XI at junction 
of R2 and R3, and a more positive emitter-biasing voltage at 
junction of R3, R4, and R19. 

Second Voltage-Divider Current (R7, R8, Rl, and R19)— 
Provides a small positive base-biasing voltage for X2 at 
junction of R7 and R8, and a more positive voltage at 
junction of R8 and Rl for the biasing cathode of M2. 

Third Voltage-Divider Current (R12, R13, and R19)— 
Provides a small positive base-biasing voltage for X4 at 
junction of R12 and R13. 

Fourth Voltage-Divider Current (R15, R16, and R19)— 
Provides small positive base-biasing voltage for X5 at 
junction of R15 and R16. 

Fifth Voltage-Divider Current (SP1, R21, and R20)— 
Provides a low positive base-biasing voltage for X7 at the 
junction of the two resistors. 

Sixth Voltage-Divider Current (SP1, R22, and R20)— 
Provides a low positive base-biasing voltage for X8 at the 
junction of the two resistors. 

124 



4. Driving current for each transistor — these currents have 
been shown in the same colors as the currents which excite 
them, (dotted blue for XI and X5; dotted red for X2, X3, 
X4, X6, X7, andX8). 

5. Oscillating currents in the five tank circuits (dotted blue in 
RF tank, dotted green in oscillator tank, and dotted red in the 
5 IF tank circuits). 

6. Diode detector current through M2 (dotted blue). 

7. Filter currents — these currents have not been shown in the 
diagrams, but they flow through the following capacitors to 
ground: C5; CI and C2B; C7, C9 and Cll; C8, C12 and C2B; 
C2A, C14. 

Details of Operation 

In any transistor, two different electron currents — the base- 
emitter current and the collector-emitter current — must flow 
during normal operation. In the PNP transistor, such as those 
employed exclusively in this particular receiver, both of these 
currents exit from the emitter terminal of the transistor. The 
base-emitter current, which is frequently referred to merely as 
emitter current, enters the base terminal and flows through the 
emitter portion of the transistor before leaving at the emitter 
terminal. 

The collector-emitter current, which is commonly referred to 
as collector current, enters the transistor at the collector terminal 
and flows through the collector, base, and emitter, within the 
transistor (in that order) before leaving via the emitter terminal. 

In NPN-type transistors, both of these flow directions are re- 
versed. The currents have the same names and they flow along 
the same two paths, except that they flow in opposite directions 
than for a PNP transistor. 

The most important physical characteristic of the transistor, 
the one which permits it to function as an amplifier, is that prop- 
erty which permits the base-emitter current to control the flow 
of collector-emitter current. A small amount of base-emitter cur- 
rent flowing will permit a large amount of collector-emitter 
current to flow, and a small change in the amount of base-emitter 
current will cause a large change in the amount of collector- 
emitter current flowing. These current changes will always have 
the same sign, or phase, meaning that an increase in the base- 
emitter current brings an increase in the collector-emitter cur- 
rent, and vice versa. 

Because of the foregoing characteristics, transistors are con- 
sidered to be current-operated devices. However, the amount of 
base-emitter current which flows at any instant is precisely de- 

125 




Fig. 7-1. Operation of a typical transistor broadcast 



126 




3HHi" 



receiver — all significant currents identified. 



127 



termined by the voltage difference existing between the base 
and the emitter. In other words, these two terminals are "biased" 
by the voltages which exist on them. It is important, therefore, 
that the reader be able to understand how these voltages are 
achieved and see what factors cause them to vary, and in what 
manner and degree. 

OPERATION WHEN NO SIGNAL IS BEING RECEIVED 

Consider the conditions which exist in the receiver when it is 
not tuned to a station. None of the signal currents (RF, IF, or 
audio) will be flowing, but all of the voltage-divider currents 
(solid blue) will be flowing. Fig. 7-2 shows these currents by 
themselves for additional clarity. Each of these currents will 
create an initial voltage difference of appropriate polarity to per- 
mit some electron current to flow from base to emitter within the 
transistor. 

XI Biasing 

The voltage at the junction of R2 and R3 is positive as a result 
of the voltage-divider current flow, and the base of XI is positive. 
However, the voltage at the lower end of R3 is even more positive, 
because it is closer to the power supply. The emitter of XI is con- 
nected to this point through R4; therefore it is more positive than 
the base, so that base-emitter current will begin flowing through 
the transistor. 

This base-emitter current (shown in solid green in Figs. 7-1 and 
7-3) flows upward through R2, through the transistor in the di- 
rection indicated, then through R4 from right to left, where it 
joins the voltage-divider current and is drawn through R19 and 
into the positive terminal of the battery. From the positive ter- 
minal of the battery, the electrons flow through the battery, out 
the negative terminal, and back to ground. 

Once a small amount of base-emitter current begins to flow, 
the collector-emitter current also begins to flow. This current 
(shown in solid red in Figs. 7-1 and 7-3) is drawn upward through 
R6, through the lower portion of the primary winding of L3, up- 
ward through the secondary winding of L2, downward through 
the collector, base, and emitter of the transistor, then through R4 
where it also joins the voltage-divider current on its journey 
to and through the power supply. 

Each of these two transistor currents will create a separate 
component of voltage drop across R4. Each voltage drop will tend 
to make the left end of R4 more positive than the right-hand 
terminal. Since the transistor base is connected through R3 to 

128 



the left end of R4, these two components of positive voltage will 
partially neutralize or nullify the original biasing condition caused 
by voltage-divider current flowing downward through R3, The 
net result will be a reduction in the amount of base-emitter cur- 
rent, and an accompanying reduction in collector-emitter current, 
from those amounts which would otherwise flow if R4 were not 
in the circuit. 

The base-emitter and the collector-emitter currents will quickly 
stabilize at values that will allow the base voltage to remain 
slightly more negative than the emitter voltage. (A tenth or two- 
tenths of a volt difference between these two terminals is typi- 
cal.) When the voltages at the base and emitter of a transistor are 
of such magnitudes that base-emitter current flows through the 
transistor, it is said to be forward-biased. If these voltages are 
such that base-emitter current cannot flow, then the transistor 
is reverse-biased. 

In addition to their usefulness in describing the total voltage 
difference between base and emitter, these terms are also used 
to describe individual components of this total voltage difference, 
or bias. When used in this sense, the voltage-divider current 
flowing downward through R3 is said to contribute a substantial 
component of forward bias to the transistor, and the two tran- 
sistor currents flowing through R4 contribute, or add, some re- 
verse bias which reduces the forward bias. In the absence of any 
received signal, each of the transistor currents will be a pure DC, 

The base-emitter and collector-emitter currents can never cut 
off the transistor entirely because, with no electron current flow- 
ing through R4, the emitter voltage would immediately rise to the 
high positive value existing on the main power-supply line. This 
would create so much forward bias that the transistor currents 
would again begin flowing. 

Similar stories about DC operating conditions can be told about 
transistors X2, X4, X5, X7, and X8. Each of these transistor cir- 
cuits is initially biased in the forward direction by the flow of one 
of the voltage-divider currents previously discussed. This initial 
forward bias starts the flow of base-emitter current which in turn 
starts the flow of collector-emitter current. In each case, the flow 
of the two transistor currents through an adjacent resistor will 
alter, or modify, the initial biasing conditions by contributing 
some reverse bias. And in each case, the amounts of the two 
transistor currents will settle down or stabilize at values which 
will permit the total bias (voltage difference between base and 
emitter) to remain at about a tenth of a volt in the forward direc- 
tion. In the PNP transistor, this means that the emitter must be 
more positive than the base. 

129 




Fig. 7-2. Operation of a typical transistor broadcast 



130 




receiver — voltage -divider and signal currents. 



131 



X6 Biasing 

The manner in which this stage is biased differs from those dis- 
cussed previously. Whereas the others are initially biased by one 
of the voltage-divider current actions, the base of X6 receives its 
operating bias directly from the emitter junction of the preceding 
stage. We have already seen how the two currents flowing through 
X5 must flow downward through emitter resistor R17, and in so 
doing how they reduce the voltage at the top of R17 from a high 
positive to a low positive value. This voltage becomes both the 
emitter voltage for X5 and the base voltage for X6. Since the 
emitter of X6 is connected to the high positive voltage, an initial 
base-emitter current, shown in solid green, will begin flowing 
through X6. This will cause the much larger collector-emitter cur- 
rent (solid red) to begin flowing. Both of these currents will flow 
downward through R18, and in so doing will reduce the positive 
voltage at the top of R18 from a high positive to a low positive 
value. These currents will stabilize or settle down at values which 
will keep the emitter of X6 only a fraction of a volt more positive 
than the base. 

Detector Current 

The final electron current which flows in this transistor receiver 
when no station or signal is tuned in is the detector current 
(dotted blue in Fig. 7-1) through M2. The detector, of course, 
will conduct electrons in only one direction — from cathode (the 
straight line in the symbol for M2) to anode (the triangle in the 
symbol). This current originates at ground below R7 and flows 
successively through R7, R8, M2, the secondary winding of L5, and 
R19 before being drawn into the positive terminal of the battery. 
This current might very easily be classed as one of the voltage- 
divider currents, and at each point along its path is a slightly 
higher positive voltage than any point preceding it, and at a lower 
positive voltage than any point following it. 

OPERATION WHEN A SIGNAL IS BEING RECEIVED 

The foregoing accounts for all of the electron currents which 
flow in the absence of a received signal. When the radio is tuned to 
a station, the five radio-frequency currents come into existence. 
The first one (shown in dotted blue in Figs. 7-1 and 7-2) flows in 
the RF tuned-tank circuit composed of CI 5, C16, and the primary 
of LI. Each half -cycle of it induces a companion current to flow 
in the secondary circuit, which includes the secondary winding 
of LI, C3, and R2. Current directions and voltage polarities have 

132 



been chosen in Fig, 7-1 so that this secondary current flows down- 
ward through R2 during this particular half-cycle, developing a 
small component of negative voltage at the top of R2 which must 
be subtracted from the normal positive base voltage of 4.2 volts. 
In a PNP transistor, a less positive voltage at the base constitutes 
"forward bias" and drives an additional amount of base-emitter 
current through the transistor. This, in turn, causes an increase 
in the amount of collector-emitter current. 

A completely separate oscillation of electrons will meanwhile 
be occurring at a higher frequency in the oscillator tank circuit. 
This tank circuit (L2, C17, CIS, and R5) is designed so that the 
oscillator frequency will always be 455 kc higher than the carrier 
frequency. Part of the oscillatory voltage is coupled via C4 to the 
emitter of XI. An instantaneous voltage polarity for this tank 
oscillation has been chosen in Fig. 7-1 so that the emitter has been 
made temporarily more positive than its normal voltage of 4.3 
volts. This also constitutes forward bias in the PNP transistor so 
that another additional component of base-emitter current is 
encouraged to flow, drawing another additional component of 
collector-emitter current through the transistor. 

This additional component of collector current must first flow 
upward through the secondary winding of L2; in so doing, it in- 
duces a feedback current to flow downward in the primary wind- 
ing. Since this feedback current is approximately in phase with 
the tank current, the oscillation will be sustained or replenished 
during each cycle of operation. 

Since the biasing conditions are being simultaneously varied by 
two separate frequencies, the collector current will be caused to 
pulsate through the external circuit at each of these two fre- 
quencies. As in the case with vacuum tube mixing and converting 
circuits, the collector current also pulsates at other frequencies, 
such as the sum and difference of the two applied frequencies. 
The tank circuit at the primary of L3 is tuned to this difference 
frequency of 455 kc so that each pulsation which occurs in the 
collector current at this frequency will surge downward through 
the lower half of the inductor and sustain or reinforce one cycle 
of the oscillation. 

The IF oscillating current in L3 and its associated capacitor is 
shown in dotted red. Current directions in Fig. 7-1 are chosen as 
downward in the primary winding, thereby inducing an upward 
current in the secondary. This secondary current delivers elec- 
trons to the base of X2, thereby increasing its forward bias during 
this half-cycle and causing a momentary increase in the flow 
of both base -emitter and collector-emitter current through the 
first IF amplifier transistor X2. 

133 




Fig. 7-3. Operation of a typical transistor broadcast 



134 




receiver — base-emitter and collector-emitter currents. 



135 



The collector-emitter current of X2 flows only through the 
upper half of the primary winding of L4. However, this provides 
sufficient coupling to the entire primary winding so that another 
IF tank current oscillation will be set up and sustained by means 
of autotransformer action. In Fig. 7-2 the upward pulsation of 
collector current through the primary winding is indicated as be- 
ing in phase with the upward flow of tank current. In the second- 
ary winding of L4, the induced current is shown as flowing down- 
ward, removing electrons from the area near the base of X3 and 
thus making the base more positive. This constitutes reverse 
bias, with the result that the amounts of base-emitter current and 
collector-emitter current through X4 will be reduced. 

The final oscillation of electrons at the intermediate frequency 
occurs in tank circuit composed of L5 and its associated capacitor. 
Each cycle of it is sustained by a single pulsation of the X4 col- 
lector current as it surges through L5. The tank voltage polarity 
shown in Fig. 7-1 corresponds to the half-cycle when a pulsation 
of collector current is not occurring. This polarity is shown re- 
versed in Fig. 7-2 along with all other tank voltage polarities and 
current directions. 

The flow of tank current up and down through the primary 
winding of L5 induces a companion current to flow down and up 
respectively through the secondary winding. This secondary 
current and the induced voltage associated with it in effect drive 
diode M2 and cause it to detect or demodulate the audio intelli- 
gence which has been carried to the antenna by the RF carrier 
signal, and which has been carried through the "front end" of the 
radio by the converter and IF amplifiers. 

Detector Current 

The detector current which flows through M2 is shown in 
dotted blue. This current flows continuously, originating at the 
ground connection below R7, flowing upward through R7 and R8, 
from cathode to anode of M2, downward through the secondary 
winding of L5 to the main power supply line of the receiver. When 
no signal is being received, this current flows as pure DC through 
the resistive portion of the path, and as pulsating DC through M2 
and L5. This feature is made possible by the integrating action of 
C12 along with R8 and R7. A positive charge will be accumulated 
on the right hand plate of C12. Electrons will flow continuously 
into this point from R8, tending to discharge it to zero. But elec- 
trons are also drawn continually away from this point to flow 
through M2 and on to the positive voltage of the power supply. 

When no signal is being received, a condition of equilibrium is 
established. The quantity of electrons flowing into C12 equals the 

136 



quantity being drawn out, and a positive voltage exists on the 
right-hand plate. When a signal is being received, the voltage at 
the upper end of the secondary winding of L5 fluctuates between 
higher and lower positive values because of the movements of the 
IF current (shown in dotted red) induced in this winding. In Fig. 
7-1, this IF current and associated voltage polarity are such that 
the upper electrode (the anode) is made less positive than it was 
before. This restricts the flow of electron current through M2. 

In the alternate half-cycle shown in Fig. 7-2, the anode of M2 
is made more positive by the IF current/voltage combination in 
the secondary winding of L5. This momentarily increases the flow 
of electrons through M2. Because of the large size of C12, the 
fluctuations in detector current are drawn directly from the right 
hand plate of C12 without causing a significant change in its total 
voltage. 

When a modulated carrier signal is being received, the strength 
of the IF oscillation in the primary of L5 varies from cycle to 
cycle. A modulation peak is characterized by a succession of 
relatively strong individual cycles of IF. A modulation trough is 
characterized by a succession of relatively weak IF cycles. One 
cycle of audio voltage consists of one modulation peak and one 
modulation trough; together they will encompass many hundreds 
or even several thousand cycles of the IF oscillation. 

The audio or modulating voltage, which is the intelligence we 
seek to hear from our radio, makes its first appearance in the radio 
on the right-hand plate of C12. The positive voltage at this point 
rises and falls at an audio rate. During modulation peaks, the 
strong IF cycles will draw an increased number of electrons away 
from C12, and the positive voltage at this point must increase. 
During modulation troughs, the weaker IF cycles will draw a 
reduced number of electrons away from C12, and the positive volt- 
age at C12 will go up again. Thus, the positive voltage on C12 
rises and falls in accordance with the strength and the frequency 
of the modulating voltage (the audio) . 

The second voltage-divider current flows downward through 
Rl. When no modulated signal is being received, this current is a 
pure DC. When an audio voltage appears on C12, this voltage- 
divider current will pulsate at an audio rate. When the voltage 
on C12 reaches its most positive value during the modulation 
troughs, it approaches more nearly the value of the voltage exist- 
ing at the bottom of Rl. Thus, there will be a reduction in the 
amount of current flowing downward through the volume control 
during modulation troughs. 

When the voltage on C12 reaches its least positive value during 
modulation peaks, it differs by a greater amount from the voltage 

137 



which exists at the bottom of Rl. Consequently, the amount of 
electron current flowing downward through Rl must increase 
during modulation peaks. 

As a result of these pulsations in current flowing downward 
through Rl, the voltage at any point along Rl will also pulsate 
at the same audio frequency. This pulsating voltage is coupled to 
C13, where it drives a small component of current up and down 
through R15. R15 functions as a base-driving resistor in much 
the same manner that a grid driving resistor functions in a vacuum 
tube circuit. During modulation peaks when the voltage divider 
current flowing downward through Rl increases, electrons will 
flow onto the left hand plate of C13, driving an equal number out 
of the right hand plate and downward through R15. This action 
develops a small component of negative voltage at the top of R15 
which must be subtracted from the positive voltage developed at 
that point by the upward flow of the fourth voltage-divider cur- 
rent and the base-emitter current through this same resistor. 

The reduction in positive voltage at the base of X5 constitutes 
forward bias in the PNP transistor, with the result that both the 
base-emitter and the collector-emitter currents through X5 will 
increase. This constitutes a half -cycle of the audio signal. The base 
driving current has been shown in dotted blue. It flows downward 
in Fig. 7-1. 

Fig. 7-2 shows the base-driving current flowing upward through 
R15. This action constitutes the half of the audio cycle correspond- 
ing to a modulation trough. When it flows upward through R15, it 
adds an additional component of positive voltage to the other 
positive voltages already there. An increase in the positive volt- 
age at the base of a PNP transistor constitutes reverse bias, so that 
both the base-emitter and collector-emitter currents will decrease 
during this half-cycle. 

During the modulation peak, the increase in transistor currents 
through X5 will develop an increased voltage drop across R17, 
which can be observed as a decrease in the positive voltage exist- 
ing at the emitter of X5, the base of X6, and the upper end of R17. 
Thus, an additional component of forward bias is applied to the 
base of X6 so that its two transistor currents also increase during 
a modulation peak. 

Push-Pull Output Amplifier 

The two output transistors (X7 and X8) are connected in push- 
pull arrangement to provide a heavy audio current to drive 
speaker SP1. Operating, or bias, voltages for these transistors are 
selected so that when one of them conducts, the other one will be 
cut off. 

138 



Both transistors are driven from the secondary winding of Tl. 
The collector current for X6 pulsates upward through the primary 
winding of Tl at the audio frequency, and each pulsation induces 
a half-cycle of current to flow downward in the secondary wind- 
ing. The induced voltage associated with this induced current is 
indicated in Fig. 7-1 as positive at the top of the secondary winding 
and negative at the bottom. This applies reverse bias to the base 
of X7, making the base more positive than the emitter and cutting 
off the flow of both transistor currents through X7. 

It also applies forward bias to the base of X8, making its voltage 
significantly less positive than the 6 volts applied to the emitter, 
and causing an increase in base-emitter current, and a consequent 
heavy surge of collector-emitter current through X8. 

In the succeeding half-cycle an opposite set of conditions pre- 
vail. The collector current through X6 decreases, the voltage 
induced across the secondary winding of Tl is negative at the top 
and positive at the obttom. This cuts off both of the transistor 
currents through X8, and causes an increase in base-emitter cur- 
rent through X7, and a resultant heavy collector-emitter current 
through X7. 

The collector currents for X7 and X8 originate at the center- 
tapped ground connection of speaker SP1. Each of these currents 
flow through only half of this winding, and of course they flow 
in opposite directions. Since they flow on alternate half-cycles 
and in opposite directions through this winding, the speaker 
diaphragm will be alternately driven to the right and left at the 
frequency of the audio modulating voltage carried by the original 
carrier signal. 

The headphone jack above X6 in the diagrams can divert the 
collector current of X6 so that it will flow either through the 
phones or through the primary winding of Tl, but not through 
both. Headphones require a much smaller power than a speaker; 
therefore the pulsations in the collector current of X6 will be 
adequate to operate the phones without the additional amplifica- 
tion provided by X7 and X8. The headphone current path has 
been indicated in Fig. 7-2. The headphone current is, of course, 
identical to the collector-emitter current of X6. 

Operation of the Overload Limiter 

The overload limiter, X3, is reverse-biased during normal op- 
eration by making the base voltage (+0.3 volts) more positive 
than the emitter voltage (+0.2 volts) so that no base-emitter 
current can flow. The base voltage is developed by the upward 
flow of collector current for X2 as it passes through RIO. (This 
same voltage also biases the collector of X2.) 

139 



The overload limiter is a form of noise limiter and might also 
be classed as an "instantaneous automatic volume control" which 
is capable of responding to a single strong noise pulse, or a very 
sudden increase in carrier signal strength. Consider the various 
circuit actions which accompany the arrival of an excessively 
strong cycle or noise pulse. 

The oscillation of electrons which goes on in L4 is sustained 
once each cycle by a pulse of collector current being drawn 
through R6 and the primary of L3 on its way to the collector of 
XI. A strong pulse will set up an excessively strong cycle of oscil- 
lation in the tank, and this will be amplified by X2 so that an 
even larger oscillation will occur in its collector tank. A portion of 
this tank voltage is coupled to the base of X3; when the instan- 
taneous tank voltage at the point where the primary winding of 
L4 is tapped is negative, the positive voltage at the base of X3 
may be reduced to the point where it is less positive than the 
emitter voltage so that X3 begins conducting. 

This conduction process in X3 will last for only a small fraction 
of a cycle, but the collector current of X3 must flow from the 
emitter to the bottom of L3. In doing so, it will deliver electrons to 
the L3 tank circuit. In order for the necessary limiting action to 
take place, this delivery of current must occur when the instan- 
taneous voltage at the bottom of the tank has been made positive 
by the oscillation of electrons. When this condition is met, the 
strength of the oscillation in L3 will be reduced for that par- 
ticular cycle. Thus, all subsequent derivations of this IF tank 
current will also be reduced in strength, and in this sense tran- 
sistors X2 and X4 will be protected against "overloading" by 
too strong a signal. 

Fig. 7-2 indicates the voltage polarities which must exist in the 
tank circuits in order for this form of degenerative feedback to 
occur. If one of these polarities should become reversed, then the 
electrons from the collector current of X3 would arrive at the 
bottom of L3 at the wrong time when the bottom of the tank is al- 
ready negative as a result of the tank oscillation. The oscillation 
would be reinforced instead of reduced in strength. 

Automatic Volume Control 

The AVC circuit of this radio is almost identical in its operation 
and construction to the positive AVC circuit discussed in Chapter 
2. Therefore the circuit actions will not be discussed again here. 
CI is the AVC storage capacitor, and R8 is the principal AVC 
filter resistor. This filter combination will react appropriately to 
reduce the signal strength when there is a long succession of 
strong carrier cycles. The overload limiter could theoretically 

140 



perform this function by reducing the strength of each one of 
these cycles. However, this would be highly undesirable on a 
continuing basis, because when an individual cycle is reduced 
in strength in this manner, its original sinusoidal wave shape is 
highly distorted. The audio intelligence carried by the modulation 
envelope of a wave shape whose every cycle required an arbitrary 
and probably varying reduction in size would surely suffer from 
such an arrangement. 

VOLTAGE AND RESISTANCE CHECKS 

The foregoing accounts for all of the electron currents which 
flow in the receiver in the absence of a received signal. Counting 
the 6 voltage-divider currents, the 14 transistor currents, and 
the detector current, it comes to 21 currents in all. In the two re- 
pair processes, known as voltage checking and resistance or 
continuity checking, these are the currents we are trying to look 
at. If any single one of these currents is significantly altered in 
quantity from its normal rate of flow, that fact can be discovered 
by the very simple maintenance process known as voltage check- 
ing. A significant variation of any one of these currents can 
occur in either direction. A current can be interrupted entirely, 
or it can be flowing in a much greater amount than is intended 
by the circuit design. Any such change in a particular current 
will inevitably be accompanied by a change in voltage at one 
or more points along the current path. When this current path 
is clearly understood, and when the current itself can be visual- 
ized, any change in an operating voltage can be instantly trans- 
lated, or interpreted, in terms of a change in current flow of a 
particular current. Once this interpretation has been made, a 
closer look can be taken along the entire path of that current, 
and the faulty component can be quickly isolated. 

Fig. 7-2 shows the operating voltage at each element of each 
transistor. For example, a base voltage of +4.2 volts, an emitter 
voltage of +4.3 volts, and a collector voltage of +.2 volts is shown 
for XI. How these voltages are achieved was discussed previously 
and will not be repeated here. 

Suppose that the radio is inoperative, or dead, due to unknown 
causes. Where and how do we start to isolate the trouble? Volt- 
age checking at various key points with a DC voltmeter set on 
an appropriate low scale is the simplest and most obvious ap- 
proach. With the common ground probe of the voltmeter con- 
nected to the ground terminal, the positive probe of the volt- 
meter should be touched to the positive terminal of the battery. 
Ml. If the voltmeter indicates that Ml is delivering its full rated 
voltage of 6 volts, the positive probe should be moved to the 

141 



left hand terminal of R19. Since all of the voltage divider cur- 
rents, all of the transistor currents, and the detector currents flow 
through R19 from left to right, a voltage drop must be generated 
across R19 in accordance with Ohm's law. This voltage drop is 
subtracted from the 6-volt output voltage of Ml so the voltage 
at the left terminal of R19 will be slightly less than 6 volts. 

If this voltage should read zero, it could mean one of two 
things, either R19 has burned out or is otherwise open, or that 
the main power-supply line has become grounded somewhere 
within the set. The quickest way to determine which one of these 
two conditions may exist is to momentarily place another resistor 
(of approximately equal value, but this is not important) , across 
the terminals of R19. If the voltage at the left terminal of R19 
then jumps up, it is a clear indication that R19 has opened, and 
should be replaced. 

If the voltage at the left terminal of R19 remains at zero, the 
main power-supply line has become grounded somewhere. This 
could have happened within any filter capacitor which connects 
the line directly to ground. Inspection of the circuit diagram in- 
dicates that C2B is the only capacitor which meets this quali- 
fication. Disconnect one end of C2B from the circuit; if the voltage 
on the line jumps up, C2B is shorted and should be replaced. 

Assume, however, that the voltmeter indicates the proper volt- 
age at the left terminal of R19. This tells us that most of the 
divider and transistor currents are flowing normally. Further, 
voltmeter checking can indicate exactly which ones are and which 
ones are not flowing. The positive probe of the voltmeter should 
now be placed on the base of XI, where it is expected to indicate 
+4.2 volts. If the voltmeter reading is zero, the voltage-divider 
current is not flowing through R2 and R3. Since resistors are 
more apt to open than to become short-circuited, each one should 
be considered in turn for the possibility that it may have failed 
and become an open circuit. 

If R3 has opened internally, the voltage-divider current can- 
not flow through it, and the transistor base will be connected 
directly to ground through R2. This could account for the very 
symptom we are confronted with — zero voltage instead of +4.2 
volts at the base of XI. A visual inspection and perhaps a con- 
tinuity check of R3 is now in order to determine its condition. 

Suppose now that R2 has burned out and opened, this would 
not account for the symptom. Instead, we would find the full 
power-supply line value at the base, because the base is con- 
nected directly to the line through R3, and no current is flowing 
through R3 because it cannot get up through R2 in the first 
place. Since no current flows through R3, no voltage drop or 

142 



difference can exist between its terminals. Consequently, an in- 
dication of full power-supply voltage at the base of XI should 
lead us immediately to suspect R2 of having failed, and a visual 
and continuity check should be instituted. The continuity check 
may be preceded by the more simple test of bridging R2 tem- 
porarily with another resistor of approximately equal value. If 
the base voltage drops immediately to the desired value of +4.2 
volts, it tells us that the voltage-divider current is now flowing 
normally through the alternate resistor and R3. No further test 
is needed to ascertain that R2 is not conducting electron current. 
In other words, it has failed and should be replaced. 

There is one other possible cause for the base of transistor XI 
being at zero voltage. C3, which couples the input carrier signal 
from the antenna tank circuit to the base of XI, may have become 
internally shorted, so that the base of XI is connected directly 
to ground through the low resistance of the secondary winding 
of LI. To eliminate this as a possibility, the coupling connection 
from C3 to the base must be temporarily opened, probably by 
unsoldering one of the capacitor terminals. If this action results 
in a normal base voltage reading, we know that voltage divider 
current is now flowing normally through both R2 and R3. No 
further tests should be necessary to tell us that C3 has failed 
and should be replaced. 

Suppose, however, that the meter reading indicates that the 
normal base voltage of +4.2 volts exists. The positive probe of 
the voltmeter should now be moved to the emitter terminal of 
XI, where we hope and expect to find +4.3 volts. If, instead, zero 
voltage is found, it can mean one of two things. Either R4, which 
connects the emitter to the positive line, has opened, or C4 has 
shorted so that the emitter is connected to ground through the 
low resistance primary winding of L2. A resistor of value equal 
to R4 may be bridged across its terminals-. If this action results 
in a normal emitter reading of +4.3 volts, we know that the two 
transistor currents are now flowing normally through the alter- 
nate resistor and they were not flowing through R4. The usual 
visual and continuity check of R4 should be made; there can be 
little doubt of their outcome (assuming that good solder joints 
existed at its terminals) . 

To determine whether C4 is internally shorted it will be neces- 
sary to open the coupling connection from the oscillator tank 
by unsoldering one of the terminals of C4. If this action results in 
a normal meter reading for the emitter voltage, C4 is the faulty 
component and should be replaced. 

If our initial emitter voltage reading has a positive value, 
greater than the desired value of +4.3 volts, a different group 

143 



of failure possibilities are implied. If the emitter exhibits the 
full power-supply voltage, the most likely reason is that neither 
of the two transistor currents are flowing through R4. Therefore 
they cannot generate the expected voltage drop across it, and the 
resistor's two terminals must be at the same voltage. Outright 
failure of the transistor to operate may be the reason for this 
symptom; however, you should first ascertain that the collector 
portion of the circuit is functioning properly. R6 may have opened, 
and since the collector current for XI must flow through R6 be- 
fore it reaches the collector and R4, it might be the cause of the 
trouble. If this were the case, the base-emitter current could 
still be flowing through the transistor and R4. The base-emitter 
current may be only one-fiftieth or so as large as the collector 
current, consequently, when flowing by itself, the voltage drop 
across R4 would be so small it would be hard to detect. 

The next step is to move the positive voltmeter probe to the 
collector terminal of XI, where the schematic tells us to expect 
voltage of about +0.2 volts. This positive voltage is generated 
at the upper terminal of R6 by the upward flow of collector- 
emitter current through it. A zero voltage reading at the collector 
would most likely indicate that an unintentional ground had 
occurred, perhaps within G5 or between the primary and sec- 
ondary windings of L2 or L3. The capacitor should be checked 
first by unsoldering one of its terminals and noting if the col- 
lector voltage returns to normal. If it does not, the trouble lies 
elsewhere, and a more detailed continuity check with an ohm- 
meter should be made between the various terminals of L2 and 
also L3. 

If R6 has opened, the voltmeter will probably indicate some 
indeterminate value of positive voltage, but higher than the 
normal 0.2 volts, at the collector. This would in fact be the posi- 
tive voltage which was stored on the upper plate of C5 at the in- 
stant that R6 failed. This stored voltage may actually increase 
slightly after the resistor fails, as a few additional electrons are 
drawn out of the upper plate of C5, to pass through the transistor 
as a tiny component of collector-emitter current. 

If failure of R6 is suspected, the usual simple expedient of 
bridging it with another resistor of equal value can be used to 
verify the fact. If transistor voltages return to normal, R6 should 
be replaced. If bridging R6 does not isolate the trouble, check 
the windings of L2 and L3 for possible open circuits. 

Voltage Checking Other Stages 

A similar procedure and analysis may be followed to locate 
faulty components associated with any of the other stages in the 

144 



receiver The voltages at each terminal of each transistor are 
given in Figs. 7-1, 7-2, and 7-3. The positive voltmeter probe 
should be touched in turn to each of the terminals of the transis- 
tors, and any abnormal reading can then be interpreted in rela- 
tion to the current or currents which bring this abnormal read- 
ing into existence. Once we recognize which of the currents has 
changed from its normal or expected value, only elementary de- 
ductive reasoning, is required to tell why this current change 
must have occurred. This leads us very quickly to the failed 
component. 

SIGNAL SUBSTITUTION TESTS 

Signal substitution is one of the two dynamic service methods, 
the other one being signal tracing. The term "dynamic checking 
means to check the receiver under complete operating conditions. 
This method should not be resorted to in the case of a faulty 
receiver until the much simpler method of voltage checking has 

been tricci* 

The signal-substitution method requires the use of a standard 
signal generator. At different points throughout the receiver 
there are several currents and voltages which are called signals. 
For example, the carrier current induced in the antenna circuit 
by the passing radio wave is universally identified as the signal. 
This current may be flowing at any frequency within the broad- 
cast band. After this carrier signal has been mixed with the 
local oscillator frequency, an IF current at a fixed frequency 
is brought into existence and amplified through the two IF 
amplifier stages. This fixed frequency is almost always 455 kc, 
and it is not at all uncommon for these IF tank voltages and cur- 
rents to be referred to as the IF signal, or some derivation of 

that term. _ 

After the IF signal has been detected, or demodulated, by M^, 
an audio current and voltage are brought into existence. These 
are subsequently amplified in the two audio amplifiers and the 
push-pull output circuit. To these several audio amplifier circuits, 
the audio voltage and current become the audio signal. 

A signal generator must be able to develop appropriate signals 
at each of these frequencies. Further, the signals so developed 
at the IF and carrier frequencies must be modulated with an 
audio signal for reasons which will be made clear. Finally, the 
signal generator must be able to develop a signal at the highest 
oscillator frequency which may be encountered— 455 kc higher 
than the upper end of the broadcast band, which extends from 
550 to about 1,600 kc. 

145 



If the standard voltage checking procedure described previ- 
ously has failed to reveal any faulty components, signal substitu- 
tion may be used to reveal which of the amplifier stages are 
functioning properly. The procedure is to start at the output end 
of the receiver (the speaker) and apply the type of signal which 
is normally expected to exist at that point to each point. For 
instance, a strong audio signal is expected to exist at either of 
the speaker primary windings. Therefore the signal generator 
should be adjusted to provide a strong audio output signal, and 
the output probe from the signal generator should be applied 
to either end of this primary winding. This will cause an audio 
current to flow up and down in the primary winding. If the 
speaker is functioning properly, the applied audio signal will 
be heard through it. If it is not heard, of course, the speaker 
itself has failed, and repair or replacement is indicated. 

If the speaker is functioning, the signal generator probe should 
be moved to either end of the secondary winding of Tl. This will 
cause audio current to flow up and down through this winding, 
and the resultant audio voltage will alternately drive the bases 
of X7 and X8, causing them to conduct. If these stages are both 
functioning properly, a much-amplified audio signal will be heard 
from the speaker. (This should remind us to turn down the 
volume on the signal generator as we move back through the 
amplifiers.) If no signal is heard, then the trouble is localized to 
either X7 or X8. 

If the previous test was satisfactory, the push-pull amplifiers 
may be certified as all right, and the probe should be moved to 
the base of driver transistor X6 (after again turning down the 
generator output) . If the audio signal is heard from the speaker, 
X6 is functioning properly, and the probe should be moved to 
the base of X5. The audio signal from the probe will vary the 
bias voltage at the base of X5 and cause the desired fluctuations 
of base-emitter current and collector-emitter current which con- 
stitute normal operation of the transistor. If this occurs, the audio 
signal will again be heard from the speaker, and in relatively few 
minutes we will have completed our check-out of the entire 
audio section of the receiver. 

Next the detector M2 should be checked for normal operation. 
The selector switch on the signal generator should now be set 
to deliver an intermediate frequency of 455 kc modulated at an 
audio frequency which is usually 400 cycles per second. This 
selector switch position may be labeled MCW, which stands for 
modulated continuous wave. 

The generator probe should now be placed at the upper end 
of the secondary winding of L5. The IF output of the signal gen- 



146 



erator will drive current up and down through this winding, alter- 
nately making the top of the winding more positive and less posi- 
tive This winding is connected directly to the power-supply 
winding line at the bottom of the diagram, which places a perma- 
nent positive voltage on the anode of M2 so that the detector 
current shown in dotted blue flows continuously. When the 
anode is made more positive by the signal generator voltage, an 
additional pulsation of detector current will flow. Thus the signal 
from the generator performs the same function as is performed 
by an IF signal coupled from L5 when the set is operating nor- 
mally. If the M2 is operating when this test is made it will de- 
modulate the 400-cycle audio signal from the artificial IF signal, 
and a 400-cycle note will be amplified through the four audio am- 
plifiers and will be heard from the speaker. 

The generator probe should next be applied to the base of the 
second IF amplifier X4. The selector switch should remain in 
the MCW position so that a modulated IF voltage is produced. 
This voltage will alternately raise and lower the positive voltage 
at the base of X4, in exactly the same manner as a normal I* 
signal would. These changes in bias voltage will cause the two 
currents flowing through X4 to fluctuate at the intermediate fre- 
quency. If this stage is operating properly, the oscillating tank 
current shown in dotted red will be set up and sustained m the 
tank circuit composed of the primary of L5 and its associated 
capacitor. The strength of this oscillation will be varied at the 
400-cycle rate in accordance with the peaks and troughs of modu- 
lation, and since we have already established that all following 
detector and amplifier stages are functioning properly, the 400- 
cycle note will be demodulated, and heard from the speaker. 

The test for X2 is performed exactly as it was for X4, after 
first readjusting downward the output level of the signal genera- 
tor The IF oscillating current in the collector tank circuit (shown 
in dotted red) will come into existence if the entire X2 amplifier 
stage is operating properly, all following amplifier stages will be 
activated, and the 400-cycle note will be heard from the speaker. 
When the converter stage is tested, the possibility that either 
one of the two tuned circuits may have become inoperative must 
be recognized. We may make an initial assumption that the oscil- 
lator tank is operative, and then attempt to apply a carrier sig- 
nal to the base of XI. (The frequency setting of the generator 
should be set as close as possible to the frequency indicated by 
the tuning dial on the radio.) It may be necessary to tune the 
signal generator frequency slowly back and forth in this range 
to find the exact frequency which will mix with the oscillator fre- 
quency and produce the 455-kc IF. 



147 



If a 400-cycle note comes from the speaker, it indicates that the 
oscillator tank, transistor XI, and the tank circuit composed of 
L3 and its associated capacitor are functioning properly. This 
leaves the antenna tank circuit and the coupling mechanism from 
this tank, namely, capacitor C3 and the secondary winding of LI 
as possible failure items. The generator probe should be moved 
first to the left hand terminal of C3, and if the signal is still heard, 
we know that C3 has not failed. Then the probe should be placed 
at the top of the primary of LI so that it excites an oscillation 
of electrons in this tank. This oscillation would correspond to the 
current shown in dotted blue in Figs. 7-1 and 7-2. If the 400-cycle 
note is still heard, then this tuned circuit and the transformer 
coupling across LI may be considered satisfactory. 

If no output comes from the speaker when the artificial signal 
is applied to the base of XI, it can mean one of two things; either 
XI is not conducting electrons, or the oscillator tank is failing 
to oscillate at its assigned frequency. It is not likely that both of 
these failures will have occurred simultaneously, so let's assume 
that the oscillator tank has failed. This assumption can quickly 
be validated by setting the signal generator to a frequency 455 kc 
higher than the setting on the radio tuning dial, and applying the 
generator probe to the top of the primary of L2. If still no signal 
is heard, one of the components in this tank has probably failed. 
The tank capacitor may be shorted, or the inductor winding may 
have an open circuit. In making this test, it will probably be nec- 
essary to rock the tuning dial of the signal generator slowly back 
and forth. 



REVIEW QUESTIONS 

1. Name and describe the seven citing each component through 
principal currents in the transis- which it flows. Also state what 
torized receiver m Fig. 7-1. function the current performs. 

2. Which of the two main currents c „ . . , 

through a transistor controls or *• Jix P lain how an audio voltage is 

regulates the other transistor njade to appear on the right hand 

current and how is the amount of plate of C12 in Fi «- 7 * 1 - 

f»fL? COntr0,,ing Current regu " 7. Trace the paths of collector-emit- 

latea? ter currents through X7 and X8. 

3. Describe the complete path of Are they direct or pulsating, and 
the oscillator-emitter current if so, at what frequency? 
through transistor XI. _ _, . . x . , 

4. How is the biasing voltage at the 8 * HuSriS! fJUff"" 1 £J?k °T er " 
has* nf translator Y« «Kt«s«Jj5 load 1,ln, te r »m describe the elec- 
base of transistor X6 obtained? tron currenta that make it j. 

5. Trace the path of the detector ble for this function to be accom- 
current (shown in dotted blue), plished. 



148 



Chapter 8 

TUNED RADIO-FREQUENCY 
RECEIVER 



The tuned radio-frequency (TRF) receiver differs from the vast 
majority of receivers in use today in that it does not utilize the 
heterodyne principle. As is pointed out in Chapter 1, the heter- 
odyne principle is a frequency-changing process, in which two 
signals are "beat" against each other to obtain a new third fre- 
quency, called the "intermediate frequency." The manifold ad- 
vantages inherent in this process have led to its adoption in vir- 
tually all receiver functions— AM and FM broadcast receivers, 
TV and radar receivers, and many communications and special 
purpose receivers. 

The principal advantage of generating a fixed value of inter- 
mediate frequency in a heterodyne receiver is that all IF ampli- 
fier stages can be fixed-tuned, rather than variable-tuned. This 
enables each such amplifier circuit to be engineered for peak 
performance at the chosen fixed frequency, with little opportunity 
for or possibility of maladjustment by an operator. 

TYPICAL TRF RECEIVER 

The TRF receiver found its greatest popularity in the early 
days of radio, before frequency-converting circuits or principles 
were highly developed. Figs. 8-1 and 8-2 show a typical TRF re- 
ceiver circuit. There are three tuned circuits in the receiver, one 
being connected to the control grid of each of the first three am- 

149 



plifier tubes, VI, V2, and V3. The tuning elements, usually one 
capacitor from each tank, must be mechanically connected to- 
gether, or ganged, so that when one circuit is tuned to a new fre- 
quency, the other two tank circuits will also be tuned to the same 
frequency. 

The third amplifier, V3, utilizes the principle of grid-leak de- 
tection to demodulate the audio signal directly from the RF car- 
rier signal. From this point on the audio section of the receiver 
functions exactly as a comparable audio section in a conventional 
heterodyne receiver. 

Fig. 8-1 differs from 8-2 in that a series diode limiting circuit, 
constructed around diode V5, has been added to Fig. 8-2 to illus- 
trate a typical application of the noise limiting function. 

Identification of Components 

The individual circuit components and their principal functions 
are as follows: 

Rl, R4, R9, R12, R14 (Fig. 8-2) —Cathode-biasing resistors. 
R2, R5, R13 — Screen-grid voltage-dropping resistors. 
R3, R6 — Power-supply decoupling resistors. 
R7 — Grid-leak biasing and driving resistor for V3. 
R8 — AVC resistor. 
RIO— Plate-load resistor for V3. 
Rll — Volume control potentiometer. 
R15 (Fig. 8-2) — Noise-limiter control potentiometer. 
R16 (Fig. 8-2) — <Jrid-driving resistor for V4. 
CI, C2 — Tuning capacitors for first RF tank. 
C3, C8, C17 — Cathode-bypass capacitors. 
C4, C9, C16 — Screen-grid filter capacitors. 
C5, CIO — Power-supply decoupling capacitors. 
C6, C7 — Tuning capacitors for second RF tank. 
Cll — Tuning capacitor for third RF tank. 
C12, C14, C18 (Fig. 8-2), C19 (Fig. 8-2)— Coupling and block- 
ing capacitors. 
C13 — Plate RF bypass capacitor for V3. 
C15 — AVC storage capacitor. 
LI (Fig. 8-2) — Radio-frequency choke. 
Tl, T2, T3 — Radio-frequency transformers. 
T4 — Audio output transformer. 
VI, V2 — Radio-frequency amplifier tubes. 
V3 — Grid-leak detector and audio-amplifier tube. 
V4 — Audio power-amplifier tube. 
V5 (Fig. 8-2) — Diode noise-limiter tube. 
Ml — Power supply. 

150 



Identification of Currents 

The several "families" of currents which flow in the TRF re- 
ceiver all have familiar counterparts in the superhet receiver 
discussed in Chapter 6. These current families are: 

1. Cathode heating current (not shown). 

2. Three RF tank currents (solid blue) . 

3. Five tube plate currents (solid red). 

4 Three screen-grid currents (also in solid red) . 

5. Three cathode filter currents (two in dotted blue; one in 

dotted green). 

6. Three screen-grid filter currents (two in dotted blue; one 

in dotted green). 

7. Two power-supply filter currents (also m dotted blue). 

8. One plate-filtering current (also in dotted blue). 

9. Three audio-signal currents (solid green). 
10. AVC current (dotted green). 

Details of Operation 

The cathode heating currents are not shown in Fig. 8-1 and 8-2 
It is common practice to omit the filament windings in circuit 
diagram, because the filament circuit is isolated both electrically 
and functionally from the remainder of the circuits in a radio. 
It is universally taken for granted that tube cathodes must be 
heated before the tubes can perform their normal function ol 
conducting electrons. iU , 

Each of the three RF tank currents flows up and down through 
the secondary winding of its respective RF transformers (Tl, T2, 
or T3) • each is sustained in oscillation by the RF current flowing 
in the 'associated primary winding. Note that unlike the trans- 
formers employed in superheterodyne receivers, the primaries ot 
the coupling transformers are not tuned by a capacitor. In the 
case of Tl, an RF alternating current flows back and forth from 
the antenna to the primary winding. In the case of T2 and T3, 
the primary winding currents are pulsating direct currents, since 
they are the plate currents of VI and V2, respectively, and these 
pulsations occur at the radio frequency being received. 

All of the amplifier tubes have been biased with grid and cath- 
ode voltages so that the tubes will operate under Class-A condi- 
tions, which means that each tube conducts electrons continu- 
ously throughout an entire cycle of RF or audio voltage Each 
plate current (shown in solid red) starts at the ground^below 
the cathode and flows up through the cathode resistor, through 
the tube from cathode to plate, out the plate and through the 

151 



c 1 i -Ht * *T »L |* «Xw— # 




^vk 











Fig. 8-1. Operation of a typical TRF receiver. 



152 




Fig. 8-2. Operation of a typical TRF receiver-currents reversed from 
those in Fig. 8-1 and a noise limiter added. 



153 



plate load to the positive terminal of the power supply, then 
through the power supply to ground. 

The three pentode tubes (VI, V2, and V4) also have screen- 
grid currents. These currents, which have also been shown in solid 
red, flow out the screen-grid terminal and through the screen- 
grid resistor where they join the plate currents and flow through 
the B+ line to the positive terminal of the power supply. 

The plate currents through VI and V2 pulsate at the radio- 
frequency being received, while the plate currents through V3 and 
V4 pulsate at the frequency of the audio intelligence which is 
carried by the carrier signal. The current through V5 flows con- 
tinuously as long as the incoming signal strength is not made 
excessive by unwanted noise pulses. When the diode current is 
flowing, it also pulsates at the audio frequency being demodulated 
from the carrier. 

There are nine separate filter currents in the circuit of Figs. 8-1 
and 8-2. Filter currents have been explained many times previ- 
ously, so they will not be repeated here. Each pulsates back and 
forth, alternately storing and drawing electrons away from the 
top plate according to the needs of the circuit to which it is con- 
nected. A corresponding electron current flows up from ground 
to the lower capacitor plate or from the capacitor to ground, in 
step with the electron flow to or from the upper plate. 

The first two cathode filter currents flow at the RF rate in and 
out of C3 and C8 (shown in dotted blue) . The third cathode filter 
current (shown in dotted green) is at the audio rate, and flows 
in and out of C17. Likewise, the first two screen grid filter cur- 
rents (also shown in dotted blue) are RF currents and flow in 
and out of C4 and C9. The other screen-grid filter current flows 
at the audio rate in and out of C16 (dotted green) . 

The two power-supply decoupling currents are shown in dotted 
blue and flow in and out of C5 and CIO to prevent RF variations 
from existing on the B+ line. Another filtering current flows in 
and out of C13 in the plate circuit of V3 to remove the RF pulsa- 
tions following detection. 

The three audio-signal currents (shown in solid green) carry 
the audio signal from the point of demodulation on C12 to the 
speaker. These are the grid leakage current from V3 to C12, 
which pulsates downward through R7, the two-way audio current 
which flows up and down through Rll, and the two-way audio 
current flowing back and forth through the closed speaker circuit. 

The AVC current, shown in dotted green, flows back and forth 
through R8 at the basic audio frequency being demodulated. The 
amount of this current which flows during a single half cycle is 
very slight, because R8 has a very high resistive value. It is this 



154 



current which delivers electrons to C15, and thereby builds up 
the negative AVC voltage. The AVC current can be looked upon 
as an "equalizing" current, since it attempts constantly to equalize 
the stored voltages at the opposite ends of R8. These voltages are 
the instantaneous audio voltage on the right hand plate of C14 
and the AVC voltage on C15. When the voltage on C12 is more 
negative than that on C15 (during the modulation peaks), the 
AVC current flows downward through R8. When the voltage on 
C12 is less negative than that on C15 (during a modulation 
trough) , the AVC current flows upward through R8 

A little reflection leads to the conclusion that the AVC voltage 
on C15 will always tend to stabilize at the average voltage exist- 
ing on C12. In other words, this voltage will be midway between 
the trough and peak values. An example or two may serve to 
clarify this conclusion. First, imagine an instance where the audio 
voltage on C12 varies between -2 and -4 volts. (This is identi- 
fied as the grid leak bias voltage in Fig. 8-1 and 8-2) . This voltage 
will be (-2) volts during a modulation trough and -4 volts dur- 
ing a modulation peak a half of an audio cycle later. The AVC 
voltage will tend to assume the average value of these two volt- 
ages, or -3 volts. This voltage will be applied directly to the con- 
trol grids of VI and V2 as part of their over-all 'bias voltages. 
Now imagine that the signal strength increases due to some 
peculiar atmospheric condition. The three RF tank currents will 
all be proportionately increased in strength, and the amount ot 
grid-leak detector current flowing out of V3 each cycle will also 
be increased. As a result, the electron accumulation on the right 
hand plate of C12 will be proportionately increased, so that the 
new trough and peak voltages will now be -3 and -6 i volte, re- 
spectively. The average of these values is -45 volts; this is the 
amount of voltage which will build up on C15 under the new 
conditions. This increased negative voltage applied to the control 
grids of VI and V2 will reduce their over-all gains, and will largely 
compensate for the unwanted increase in signal strength. 

Noise-Limiting Diode Operation 

V5 is connected as a series noise limiter, and functions in sub- 
stantially the same manner as described in Chapter 3. R15 acts 
a voltage divider to "bias" the plate of V5 more positively than 
the cathode. This biasing action is accomplished by the voltage 
divider current shown in dotted red in Fig. 8-2. Because the point 
midway along R15 where the cathode voltage is tapped off will 
always be more negative than the left-hand terminal of R15, the 
diode plate current shown in solid red will flow continuously 
along the indicated path unless interrupted by some other action. 

155 



During the modulation peaks and troughs which characterize 
all audio voltages, the amount of diode current through V5 will be 
mo-eased and decreased. These variations in current flow through 
R14 will cause the positive voltage at the upper end of the resistor 
Zu*o?T ^ d + decreas f* ** same audio frequency. These audio 
voltage fluctuations will be coupled across C19 to R16 the grid 
driving resistor for V4. ' 8 

When diode current through R14 and V5 increases during mod- 
ulation peaks electron current will be drawn upward through 
R16, making the control grid of V4 positive. When the diode cur- 
rent decreases during modulation troughs, electron current will 
ne ativT WDWard throu g h R16 > making the control grid of V4 

Diode V5 can be cut off entirely only by an excessively strong 
noise pulse having a negative polarity when it reaches the plate 

*X a A? Uhe makes the plate of V5 more negative than the 
cathode, and the upward current flow through R14 is stopped The 
positive voltage at the cathode then decreases to the same value 
existing at the tap on R15. This drop in positive voltage at the 

Jwi r e h en Ri fi of R t 4 d i ves *j*y electron current d °— s 

^♦k- + u ' making the gnd of V4 suffi ciently negative to cut 
JTk I entU * ely Snd n ° ise hmithl ^ has ^ accomplished. 
With the exception of the circuits just discussed-detector, 
AVC, and noise Wer— the circuits in the TRF receiver are am- 
plifier circuits which are discussed more fully in other chapters 
of this book or m Amplifier Circuit Actions, an earlier volume in 
this series. 



1. In the TRF receiver of Figs. 8-1 
and 8-2, describe the movements 
and complete paths of those cur- 
rents which drive or sustain the 
electron currents oscillating in 
the three grid tank circuits. 

2. At what point in this receiver 
does the audio voltage become 
clearly identifiable as a separate 
voltage, no longer being "carried" 
by the carrier frequency? 

3. Describe the movements (what 
makes it flow, and what is its 
complete path) of that electron 
current which delivers electron 
charge to (or withdraws it from) 
the upper plate of capacitor C15. 
In which direction does this cur- 
rent flow during a modulation 



REVIEW QUESTIONS 



trough? During a modulation 

|)€AK T 

4. Make a comparison between al- 
ternate half cycles of this cur- 
rent, during a period of signal 
fade. How about during a signal 
build? 

5. What is the principal function 
of the voltage divider current 
(shown in dotted red) which 
flows through R15? 

6. If the plate of diode V5 becomes 
more negative than its cathode, 
explain how a heavy pulsation of 
plate current through triode V3, 
which constitutes a negative 
noise pulse, acts in order to cut 
oft* the diode V5. 



156 



INDEX 



AGC, 46 

Amplification, power, 109 

Amplifier, audio power, 108-112 

IF, 104-108 

amplifier, push-pull, 138-139 
Amplifier operation, 42-43 
Anomalies, propagation, 45-46 
Audio power amplifier, 108-112 
Audio voltage, 38-39 
Automatic gain control, 46 
Automatic volume control; see 

AVC 
Autotransformer action, 26-27 
AVC, 140-141 

current, 28, 44 

instantaneous, 140 

operation, 43-46 

voltage, 44-46, 107-108 

positive, generation of, 46- 
60 



B 

Base-emitter current, 49, 51, 

125 
Beat-frequency oscillator, 28- 

34 
Boat note, 33-34 
BFO, 28-34 

Biasing, transistor, 128-132 
Build-up, signal, 45-46 



Capacitive reactance, 9, 11 
Carrier, 7, 40 

Cathode, open, effect on volt- 
age reading, 113-114 
Circuit, audio power amplifier, 
108-112 
beat-frequency oscillator, 28- 

34 
detector, AVC, and audio 

amplifier, 35-46 
dual-diode noise limiter, 72- 

76 
filter, 85-87 

half-wave rectifier, 82-88 
IF amplifier, 104-108 
overload limiter, 139-140 
pentagrid converter, 20-28 
pentagrid mixer, 13-20 
positive AVC, 46-60 
power supply, 82-88 
regenerative detector, 89-97 
series-diode noise limiter, 65- 

72 
shunt-diode noise limiter, 61- 

65 
squelch, 77-81 
superregenerative receiver, 

97-103 
transistor broadcast receiver, 

122-148 
triode mixer, 7-12 



157 



Code reception, 29 

Collector-emitter current, 49, 
52, 125 

Constants, time-, 57-58, 86 

Converter, pentagrid, 20-28 

Coulomb's law, 41-42, 59-60 

Coupling action, 42 

Current, AVC, 28, 44 
base-emitter, 49, 51, 125 
carrier, 7, 8-11 
collector-emitter, 49, 52, 125 
detector, 132, 136-138 
feedback, 17, 26 
filament, 113 
filter, 85-88 
grid-leakage, 24 
IF filter, 37-38, 42-43 
local oscillator, 7-11, 25, 32-33 
plate, 24, 31, 42, 80, 100-101, 

105-106, 109-110 
screen-grid, 28, 32, 106-107 
signal, 7, 31 
tank, 25, 32, 100-101 
voltage divider, 49, 79-80 

Current direction, relation to 
voltage polarities, 93 

D 

Decoupling network, 18-119 
Demodulator, 35 
Detector, AVC, and audio amp- 
lifier, 35-46 
current, 132, 136-138 
operation, 37-42 
regenerative, 89-97 
disadvantages of, 89 
Difference frequency, 11-12, 

19-20 
Diode, noise limiting, 155-156 
Diode tube operation, 84-85 
Dual-diode noise limiter, 72-76 
Dynamic checks, 145 

E 

Election flow, in transistor, 48 



Fading, signal, 45 
Feedback current, 17, 26 
Filament current, 113 
Filter, circuit, 85-87 

current, 85-88 
Filtering action, 9, 11 
Formula, capacitive reactance, 
9 

power, 109 
Frequency, 19-20 

difference, 11-12 

local oscillator, 16 

sum, 11 



G 

Grid-leak bias voltage, 94, 96 
Grid-leakage current, 24 



H 

Half-wave rectifier, 82-88 
Hartley oscillator, 32 
Heterodyne receiver, advan- 
tages of, 149 



IF, obtaining of, 12, 27-28 

IF amplifier, 104-108 

IF filter current, 37-38, 42-43 

Inductance, 22 

Inertia, electrical, 22 

Intermediate frequency; see IF 



Law, Coulomb's, 41-42, 59-60 

Lenz's, 112 

Ohms, 41 
Lenz's law, 112 
Limiter, overload, 139-140 
Local oscillator current, 7 



158 



Local oscillator frequency, 16 
Long time constant, 59 

M 

MCW, 146-147 

Mixer, pentagrid, 13-20 

triode, 7-12 
Modulation, 39-40 
Morse Code reception, 29 



N 

Network, decoupling, 18-19 
Noise limiter, dual diode, 72-76 

series-diode, 65-72 

shunt diode, 61-65 
Noise limiting diode, 155-156 
Noise pulses, 64 

effect of, 68 



O 

Ohm's law, 41 

Open cathode, effect on voltage 

reading, 113-114 
Operation, AVC, 43-46 

diode tube, 84-85 

overload limiter, 139-140 
Oscillating tank voltage, 25 
Oscillations, sustaining of, 26 
Oscillator, beat-frequency, 28- 
34 

current, 7, 25, 32-33 

Hartley, 32 

quench, 97-103 

tickler coil, 16 
Output transformer, 110-111 
Overload limiter circuit, 139- 
140 



Plate current, 24, 31, 42, 80, 
100-101, 105-106, 109-110 
PNP transistor, AVC voltage 

for, 47 
Positive AVC voltage, genera- 
tion of, 46-50 
Power amplification, 109 
Power amplifier, audio, 108-112 
Power supply circuit, 82-88 
Propagation anomalies, 45-46 
Pulses 
noise, 64 
effect of, 68 
Push-pull amplifier, 138-139 

Q 

Quantity of electrons in charge, 

59-60 
Quench oscillator, 97-103 

R 

Radio, superhet, voltage check- 
ing of, 112-118 
Reactance, capacitive, 9, 11 
Receiver, 149-156 

superregenerative, 97-103 
transistor broadcast, 122-148 
signal substitution checks, 

145-148 
voltage and resistance 
checks, 142-145 
Rectifier, half-wave, 82-88 
Regenerative detector, 89-97 

disadvantages of, 89 
Requirements, signal genera- 
tor, 118 
Resistor, checking for open, 114 
Ripple voltage, 87-88 



Pentagrid converter, 20-28 
Pentagrid mixer, 13-20 



Saturation current, 76 
Screen-grid current, 28, 32, 
106-107 



159 



Shunt-diode noise limiter, 61- 

65 
Signal, normal expected, 118- 

121 
Signal build-up, 45-46 
Signal current, 7, 31 
Signal fading, 44 
Signal generator requirements, 

118 
Signal substitution, 118-121, 

145-148 
Second detector, 35 
Series-diode noise limiter, 65- 

72 
Speaker action, 111-112 
Squelch circuit, 77-81 
Sum frequency, 11 
Superhet radio, voltage check- 
ing of, 112-118 
Superregenerative receiver, 

97-103 
Sustaining oscillations, 26 



Tank current, 25-26, 32, 100-101 
Tickler-coil oscillator, 16 
Time constants, 57-58, 86 
Time periods, AVC circuit, 53 
Transformer, action, 21-24 
output, 110-111 



Transistor, biasing, 128-132 
broadcast receiver, 122-148 
signal substitution checks, 

145-148 
voltage and resistance 
checks, 142-145 
TRF receiver, 149-156 
Triode mixer, 7-12 
Tuned radio frequency re- 
ceiver, 149-156 



Voltage, audio, 38-39 
AVC, 28, 44-46, 107-108 
positive, generation of, 46- 
60 
-divider current, 49, 79-80 
grid-leak bias, 24, 94, 96 
modulation, 39 
oscillating tank, 25 
ripple, 87-88 
Voltage and resistance checks, 
transistor receiver, 142- 
145 
Voltage amplifier, 42-43 
Voltage checking, superhet 
radio, 112-118 



Zero beat, 34 



160